PSYCHE
A Journal of Entomology
Volume 68
1961
Editorial Board
Frank M. Carpenter, Editor P. J. Darlington, Jr.
W. L. Brown, Jr. H. W. Levi
E. O. Wilson H. E. Evans
Published Quarterly by the Cambridge Entomological Club Editorial Office : Biological Laboratories 1 6 Divinity Ave.
Cambridge, Mass., U. S. A.
The numbers of Psyche issued during the past year were mailed on the following dates:
Vol. 67, no. 4, Dec., 1960: July 7, 1961
Vol. 68, no. 1, March, 1961 : August 2, 1961
Vol. 68, nos. 2-3, June-Sept., 1961: December 29, 1961
PSYCHE
A JOURNAL OF ENTOMOLOGY
Established in 1874
Vol. 68
March, 1961
No. 1
CONTENTS
Australian Carabid Beetles V. Transition of Wet Forest Faunas from New Guinea to Tasmania, P. J. Darlington , Jr. 1
A Reconsideration of the Genus Epipompilus ( Hymenoptera : Pompilidae).
H. E. Evans 25
The Rediscovery and Probable Phylogenetic Position of Psilopsocus (Psocoptera) . E. L. Mockford
38
CAMBRIDGE ENTOMOLOGICAL CLUB
Officers for 1961-62
President ...J. J. T. Evans, Harvard University
Vice-President C. Walcott, Harvard University
Secretary A. R. Brady, Harvard University
Treasurer F. M. Carpenter, Harvard University
Executive Committee R. W. Taylor, Harvard University
S. K. Harris, Boston University
EDITORIAL BOARD OF PSYCHE F. M. Carpenter (Editor), Professor of Entomology , Harvard University
P. J. Darlington, Jr., Head Curator of Insects, Museum of Com- parative Zoology
W. L. Brown, Jr., Assistant Professor of Entomology, Cornell University; Associate in Entomology, Museum of Comparative Zoology
E. 0. Wilson, Associate Professor of Zoology Harvard University H. W. Levi, Associate Curator of Arachnology, Museum of Com- parative Zoology
H. E. Evans, Associate Curator of Insects, Museum of Comparative Zoology
PSYCHE is published quarterly by the Cambridge Entomological Club, the issues appearing in March, June, September and December. Subscription price, per year, payable in advance: $4.50 to Club members, $5.00 to all other subscribers. Single copies, $1.25.
Checks and remittances should be addressed to Treasurer, Cambridge Ento- mological Club, Biological Laboratories, Harvard University, Cambridge, Mass.
Orders for back volumes, missing numbers, notices of change of address, etc., should be sent to the Editorial Office of Psyche, Biological Laboratories, Har- vard University, Cambridge, Mass.
IMPORTANT NOTICE TO CONTRIBUTORS Manuscripts intended for publication should be addressed to Professor F. M. Carpenter, Biological Laboratories, Harvard University, Cambridge, Mass.
Authors contributing articles over 8 printed pages in length may be required to bear a part of the extra expense, for additional pages. This expense will be. that of typesetting only, which is about $6.00 per page. The actual cost of preparing cuts for all illustrations must be borne by contributors; the cost for full page plates from line drawings is ordinarily $12.00 each, and the full page half-tones, $18.00 each; smaller sizes in proportion.
AUTHOR’S SEPARATES
Reprints of articles may be secured by authors, if they are ordered at the time proofs are received for corrections. A statement of their cost will be furnished by the Editor on application.
The December i960 Psyche (Vol. 67, no. 4) was mailed July 7, 1961. The present issue of Psyche (Volume 68, no. 1) is pub- lished with the aid of a National Science Foundation Grant (G 15817) to the Cambridge Entomological Club.
The Lexington Press. Inc., Lexington, Massachusetts
PSYCHE
Vol. 68 March, 1961 No. 1
AUSTRALIAN CARABID BEETLES V.1 TRANSITION OF WET FOREST FAUNAS FROM NEW GUINEA TO TASMANIA
By P. J. Darlington, Jr.
Museum of Comparative Zoology, Cambridge, Mass.
Introduction
Beetles of the family Carabidae (predaceous ground beetles) are numerous in tropical rain forest in New Guinea and numerous also (but less diverse) in cool south temperate rain forest in Tasmania, but no species and hardly any genus is common to the two faunas, and even the dominant tribes are different. However there is no single boundary between the New Guinean and Tasmanian faunas, but a broad and complex transition, which I shall try to describe.
My interest in this part of the world began with the Harvard Australian Expedition of 1931-1932, when I collected Carabidae in eastern Australia north to part of the Cape York Peninsula, as well as in southwestern Australia. In 1943-1944 I spent eleven months in New Guinea as an army entomologist, and was able to collect Carabidae especially in lowland rain forest at Dobodura, Papua, while hospitalized there, and in mountain forest on the Bismarck Range, Northeast New Guinea, in lieu of leave. I have sorted and arranged my own and much borrowed material and am now more than half way through writing “The Carabid Beetles of New Guinea” (see Darling- ton 1952), so that I have a good knowledge of New Guinean Carabi- dae. Recently, from December 1956 to June 1958, I have been again in eastern Australia, traveling and living in a small truck with my wife and fourteen-year-old son, and collecting Carabidae in practically every important piece of wet forest from the northern tip of Cape
Tarlier parts of this series are listed in the reference list at the end of this paper.
]
2
Psyche
[March
York to the southern tip of Tasmania.2 A brief itinerary with maps and list of localities has been published (1961). Information and collections obtained during this trip have enabled me to correlate other information and write the present paper. New genera and species re- ferred to now (but not by name) will be described in forthcoming numbers of Psyche and Breviora.
The Forests
My “wet forests” are rain forests as classified in “The Australian Environment” (CSIRO 1950, 77-96). That is, they are dense, ever- green (non-deciduous) forests with closed canopies, often (in tropical rain forest) with many woody vines, but with comparatively little low vegetation, the ground being covered with dead leaves and leaf mold rather than grass or herbs.
Two main types of rain forest exist in the Australian Region: tropical (including subtropical) (Figs. 1, 2) and south temperate (Figs. 3, 4). Tropical rain forest is widely distributed in New Guinea at low and middle altitudes, although in the drier country of southern New Guinea it is replaced by op n savannah woodland like that of much of northern Australia. Tropical rain forest occurs also on the eastern edge of Australia in separate tracts spaced irregularly from parts of Cape York south through Queensland and northern New South Wales (map, Fig. 6). The best of this forest in tropical and subtropical Australia as well as in New Guinea is real, Malaysian- type rain forest, although some tracts in Australia are lighter and seasonally drier, and light rain forest sometimes grades into semi- deciduous monsoon forest.
The northernmost rain forest in Australia is the tip-of-peninsular (Fockerbie or Somerset) tract on the tip of Cape York. It is lowland rain forest, but somewhat depauperate (see p. 17) .
2This trip was supported in part by a fellowship of the John Simon Guggen- heim Memorial Foundation. I am especially indebted to Dr. L. J. Webb, of the Commonwealth Scientific and Industrial Research Organization, for in- formation on the distribution of rain forest in Queensland, to many members of the Queensland Department of Forestry who aided or guided us in the field, and to Mr. P. J. Killoran, of the Queensland Department of Native Affairs, who arranged our visit to Bamaga and the tip of Cape York. I very much regret that I do not have space to acknowledge other assistance in detail here.
Explanation of Plate 1
Fig. 1. Tropical rain forest, Lake Barrine, Atherton Tableland, North Queensland (P. J. D. 1932).
Fig. 2. Interior of tip-of-peninsular (tropical) rain forest, from edge of new clearing, Lockerbie, Cape York, Queensland (P. J. D. 1958). This is the habitat of Mccynognathus.
Psyche, 1961 Vol. 68, Plate
Psyche, 1961
Vol. 68, Plate 2
4
1961]
Darlington — Australian Carabid Beetles
5
Next in order southward is a gap more than ioo miles wide of drier, open savannah woodland (Fig. 5) in which may be an isolated piece of rain forest near the head of the Jardine River, unknown biologically (Brass 1953, pp. I54> 161).
Next is the mid-peninsular rain forest system. It extends irregularly and with perhaps slight interruptions from near Iron Range and Mt. Tozer south to the “Rocky Scrub” east of Coen. Altitudinally it ex- tends from near sea level ( e.g . at Iron Range) to about 2,000 ft. on the higher summits of the Mcllwraith Range. It includes fairly heavy rain forest, although its quality varies locally.
Fig. 5. Rather dry savannah woodland northeast of Coen, Cape York peninsula. (P. J. D. 1932). Such woodland is an effective barrier to rain forest Carabidae in the tropics.
Next, after another gap more than 150 miles wide of drier, open woodland, is the base-of-peninsular or main tropical rain forest system of North Queensland. Outlying pieces of semi-rain forest of this system are within sight of Cooktown, and heavier rain forest begins on the coastal mountains (Mt. Amos, Mt. Finnigan) about 20 miles to
Explanation of Plate 2
Fig. 3. South temperate rain forest, Lake St. Clair, Tasmania (P, J, D. 1957). On left is transitional wet forest with overstory of big eucalypts; center, heavy rain forest including N othofagus.
Fig. 4. Interior of old south temperate rain forest, Cradle Valley, northern Tasmania (courtesy Mr. H. J. King, Honorary Photographer, and Mr. Frank Ellis, Director, Queen Victoria Museum, Launceston).
6
Psyche
[March
the south. From here an irregular system of good rain forests extends somewhat discontinuously but with no very wide breaks south to and across the Atherton Tableland and farther south along a series of plateaus and ranges to the Mt. Spec plateau (Paluma Range) almost within sight of Townsville. Much of this forest system lies between 1,000 and 5,000 ft. altitude, but areas of good rain forest belonging to it occur (or occurred before being cleared) also on the coastal plain east of the Atherton Tableland and in the Mossman-Daintree region.
From the southern end of the main tropical rain forest system to below Rockhampton is a gap of nearly 500 miles of dry, open wood- land broken only (so far as I know) by two noteworthy islands of rain forest. One is at about 3,000-4,000 ft. on the crest of the Elliot Range, within sight of (southeast of) Townsville but separated from the northern rain forests by a low, comparatively dry valley. The other, more important island of rain forest is on the Eungella Range about 40 miles inland from Mackay, at about 2,000-4,000 ft. altitude. Scattered fragments of semi-rain forest, for example near Proserpine (Repulse Bay) and Yepoon (Byfield), are relatively unimportant so far as carabid distribution is concerned.
South of Rockhampton, in the edge of the south temperate zone, begins what I call the subtropical rain forest system. The first piece of (rather poor) rain forest of this system is on Mt. Jacob east of Many Peaks. Other tracts are widely scattered in southeastern Queensland at low altitudes as well as on mountains (Blackall Range, Bunya Mts., Mt. Tamborine, McPherson Range on the New South Wales border, etc.). The different forest tracts vary in quality, but the best of them approximate tropical rain forest. This system of rain forests extends into northeastern New South Wales at rather low altitudes, although much of it has now been cleared. The more important pieces that still remain are listed and briefly described in my published locality list (1961). The most southern good tract that seemed to me to be tropical-type rain forest is on “Mt. Dorrigo”, on the lower (eastern) edge of the Dorrigo Plateau, at about 30° 20' S., but small pockets of more or less similar forest occur still farther south, even south of Sydney, especially in wet ravines.
South temperate rain forest (see again Figs. 3, 4) is different in aspect from tropical rain forest (fewer vines, etc.) and different botanically, often dominated by southern beeches (Nothofagus) . Such forest is widespread in southwestern Tasmania and occurs in isolated tracts elsewhere in Tasmania (see paper referred to above for details). Isolated tracts of similar forest occur on plateaus and moun- tains in southern Victoria including the Otway Ranges southwest of
1961]
Darlington — Australian Carabid Beetles
7
Melbourne and some of the southern “Victorian Alps” east of Mel- bourne. This kind of forest occurs also, at wide intervals, on isolated plateaus in eastern New South Wales, notably on the plateau of the Mt. Royal Range (Barrington Tops and Tomalla Tops) at about 310 50' S. and on the higher part of the Dorrigo-Ebor Plateau (especially at Point Lookout in New England National Park) at about 30° S. Both these plateaus reach about 5000 ft. above sea level. The northernmost Nothofagus in Australia is still farther north, on the southern border of Queensland, where small tracts of old trees exist on the highest points of the McPherson Range, at about 28° 20' S. and 4,000 ft. altitude. Nothofagus does not occur on the mountains of tropical North Queensland but is dominant in New Guinea in mountain forests between about 6,500 and 10,000 ft. (Womersley and McAdam 1957, p. 25). However, south temperate groups of Carabidae do not occur in the New Guinean Nothofagus forests.
The distribution of tropical (including subtropical) and south temperate rain forest is shown, rather diagramatically, on the accom- panying map (Fig. 6). The map is based partly on the vegetation map in “The Australian Environment” (CSIRO 1950, pp. 88-89) and on Brass’s (1953, p. 152) map of Cape York rain forests, but many details are modified according to my own observations. In most cases rain forest is not continuous within the boundaries shown, but occurs as irregular, sometimes discontinuous tracts and strips inter- spersed with savannah woodland (in the north) and/or sclerophyll forest (in the south). The two kinds of rain forest overlap widely in New South Wales. Within the area of overlap south temperate rain forest is usually above (at higher altitude than) tropical rain forest, but there is some mixing.
The Carabidae
The wet-forest Carabidae of New Guinea and Australia, including Tasmania, are numerous, diverse, and complex in ecology and distri- bution. They form three general ecological groups. Those that live on the ground without being specially associated with surface water are mesophiles or geophiles. Those that live on the ground beside streams or ponds or in swamps are hydrophiles. And those that live on tree trunks or in foliage above the ground are arboreal. According to my (i943, P- 41) rough analysis of the Australian carabid fauna, at least half the species are geophiles, not quite a quarter hydrophiles, and not quite a quarter arboreal. The carabid fauna of New Guinea divides in something like the same way, although I cannot yet give exact figures.
Psyche, 1961
Vol. 68, Plate 3
NEW GUINEA
TASMANIA
1961]
Darlington — Australian Carabid Beetles
9
State of wings of Carabidae is correlated with ecology and distri- bution. Most Carabidae in most parts of the world have fully devel- oped inner wings and can fly, but some have lost their wings (except for vestiges) and become flightless. The Australian carabid fauna includes an unusually large proportion of flightless species : according to my rough analysis (loc. cit.), nearly 45% of all Australian Carabi- dae have atrophied wings, and many genera and even some tribes are wholly flightless. Most hydrophiles and arboreal forms have retained their wings and can fly, but about 75% of Australian geophile Carabi- dae are flightless, and flightless groups are common everywhere in Australia, at low and high altitudes and in wet and dry climates, and some are well represented in the tropical as well as the temperate parts of the continent. In New Guinea flightnessess is rare among lowland Carabidae. This accords with the general rule that most Carabidae in most wholly tropical lowland areas are winged. On mountains in New Guinea, however, as on many tropical mountains elsewhere, flightless geophile Carabidae are numerous.
New Guinea-Tropical Australian Relationships Probably the first fact that strikes entomologists collecting in the rain forests of tropical Australia is that some of the insects are species that occur in New Guinea. This is expected. The Australian rain forests themselves are predominantly New Guinean (or Malaysian) both in aspect and in botanical relationships (CSIRO 1950, pp. 95- 96; Brass 1953? P* 154) ; many mammals in the North Queensland rain forests belong to New Guinean genera or even species; and so do many birds. Some Carabidae are common to New Guinean and Australian rain forests. For example Syleter papua Dari, extends to the tip of Cape York, living on the ground in shaded swamps. Morion longipenne Putz. of New Guinea extends to the main North Queens- land rain forests, on and in fallen logs. And Violagonum violaceum (Chd.) is common in rain forest in New Guinea and eastern Australia south at least to near Rockhampton, in accumulations of dead leaves on the ground and in thick foliage. Besides shared species like these (there are many others among Carabidae) the New Guinean and Australian rain forests share some geographically restricted genera, for example Platycoelus ( Chlaenioidius) , Loxandrus , and Stricklandia,
Explanation of Plate 3
Fig. 6. Distribution of rain forests in eastern Australia. Solid lines enclose principal areas of tropical (including subtropical) rain forest; broken lines, of south temperate rain forest. In most cases rain forest is not continuous within the boundaries shown but occurs in discontinuous or scattered tracts. See text for further details.
10
Psyche
[March
as well as many more-widely distributed genera. Up to a point, there- fore, the Carabidae agree with the forest trees, mammals, and birds in showing a considerable number of species and genera common to the rain forests of New Guinea and tropical Australia.
When I was collecting on the Atherton Tableland in northeastern Australia in 1932, I found not only many Carabidae of obviously New Guinean groups but also, in rain forest, many species of Australian groups not known to occur in New Guinea. Included were striking endemic species of Notonomus , Trichosternus, Leiradira , Pamborus, and Mystropomus. Knowing, as I did, that the rain forests of Australia and New Guinea had much in common, and knowing that the Carabidae of New Guinea were poorly collected, I imagined in New Guinea a rich fauna of the genera just named, perhaps in rain forest at middle altitudes, but wholly unknown. It was a sort of El Dorado for the future, to a young and enthusiastic carabid student. But now that I have collected in New Guinea and seen thousands of Carabidae collected there by other persons, I know that this El Dorado does not exist, and I know why. All the Carabidae common to the New Guinean and Australian rain forests are winged and probably fly. All the genera mentioned above as represented in rain forest on the Atherton Tableland are wholly flightless, and I know now that there is no direct relationship between any flightless Carabidae of the New Guinean and Australian rain forests.3
The difference between the flightless Carabidae of Australia and New Guinea goes far beyond mere differences of species and genera. The composition and origins of the two faunas are fundamentally different. Flightless Carabidae are numerous everywhere in Australia, even at low altitudes in the tropical part of the continent including Cape York. Many of the species belong to wholly flightless genera or even flightless tribes that have evidently been in Australia a long time. Derivatives of old Australian flightless groups dominate the flightless ground-living carabid fauna of tropical rain forest in Australia. In New Guinea, in contrast, no primarily flightless groups of Carabidae occur at low altitudes. A very few species of the primarily winged
3If tiger beetles are considered Carabidae, Tricondyla aptera 01. is an exception to this rule. The genus Tricondyla is primarily Oriental and is wholly flightless. Nevertheless T. aptera has reached New Guinea, probably rather recently (it is only slightly differentiated there), and has got beyond New Guinea to the mid-peninsular rain forests of Cape York. (It has reached the Solomon Islands and New Hebrides too.) It is a good sized (nearly an inch long), big-eyed, ant-like, active insect, which lives on tree trunks in rain forest. It has probably dispersed on floating trees, which ground-living Carabidae are not likely to do.
1961]
Darlington — Australian Carabid Beetles
1 1
genera Clivina, Tachys, Lesticus , Tlatycoelus , and Loxandrus have undergone wing atrophy at low altitudes in New Guinea (Darlington in press) , but they have evidently done it recently, in situ. Some of the species are still dimorphic, with fully winged individuals occurring with the short winged ones, and all the short winged lowland species are closely related to long winged ones that still exist in New Guinea. It is only above about 5000 ft. in the mountains that flightless Carabi- dae become numerous in New Guinea, and they too have apparently undergone wing atrophy in situ. That is, they have been derived on the mountains of New Guinea from winged ancestors, and do not represent flightless stocks of other regions. This is my conclusion after making formal studies of the New Guinean representatives of the two principal tribes concerned, the Agonini (Darlington 1952, especially table p. 108) and Pterostichini (in press).
Besides the change of specific flightless stocks from New Guinea to Australia there is a change of dominance of tribes. In New Guinea, Agonini are much more numerous than Pterostichini, and most flight- less Carabidae of the island are agonines. But in Australia, even in the tropical rain forest, Pterostichini are overwhelmingly dominant and include most of the flightless forms. This striking shift of domi- nance is further discussed on page 22.
The first important finding of the present study, then, is that, al- though the rain forests of New’ Guinea and tropical Australia are similar and share many species of plants, mammals, birds, and winged insects including many winged Carabidae, they have wholly different faunas of flightless Carabidae, which differ not only in taxonomic details but also in general ecology (in relation to altitude), in origin of the flightless stocks, and in relative dominance of tribes.
Transition in Australia: South from the Tropics
Now to be considered is the transition of wet forest carabid faunas within the limits of Australia and Tasmania.
Five important genera of flightless geophile Carabidae are mentioned above as occurring in rain forest on the Atherton Tableland. Of these five genera, Notonomus is most dominant. It is a genus of about 100 species, confined to eastern and southeastern Australia and Tas- mania except for one species isolated in southwestern Australia. The genus’ northern limit is between Daintree and Cooktown. It is repre- sented by several species (some very localized) in the main tropical rain forest system of North Queensland, where it seems to be confined to rain forest. It is well represented in the subtropical rain forests of South Queensland and northern New South Wales and south through
12
Psyche
[March
eastern New South Wales and southern Victoria; in these areas some species occur not only in rain forest (including south temperate rain forest) but also in wet sclerophyll forest and good savannah woodland. However only two groups of the genus reach Tasmania and only one group (two related, primarily allopatric species) occurs in rain forest there.
Trichosternus is a genus of 25 or more species confined to eastern Australia, except that one species is isolated in southwestern Australia (Darlington 1953, p. 94). The genus’ northern limit is between Daintree and Cooktown. It occurs (several species, some very local- ized) throughout the main rain forest system of North Queensland, where it is apparently confined to rain forest. It is well represented also in the subtropical rain forest system of South Queensland and northern New South Wales, and in this area some species occur in savannah woodland as well as in tropical-type rain forest, and some have entered south temperate rain forest on the Dorrigo-Ebor plateau and the Mt. Royal Range. The southern limit of the genus is some- where in east-central New South Wales, probably not far north of Sydney.
The northern limit of Leiradira (or of the group of genera that includes Leiradira) is between Daintree and Cooktown. This genus too occurs in much of the main tropical rain forest system of North Queensland, being represented there by several distinct species each more or less localized, but the genus may be absent in the southern extension of the main tropical rain forest system south of the Atherton Tableland. It is represented also by several species in the subtropical rain forests of South Queensland etc. Its southern limit is apparently on the lower, eastern edge of the Dorrigo plateau. It is confined to eastern Australia. It is wholly or chiefly a rain forest genus in all parts of its range.
The three preceding genera are all Pterostichini. All their species are flightless geophiles. Additional flightless geophile pterostichines are localized in all the different rain forest areas of Australia from Cape York to Tasmania. Examples are Mecynognathus in the tip-of- peninsular forests; Paranurus in the mid-peninsular forests; Loxo- genius and undescribed genera in the main tropical rain forest system ; Nursus s. s.j Liopasa, Ceratoferonia Zeodera, and Notolestes in the subtropical rain forest system ; Loxodactylus in the wet forests of southern Victoria; and Rhabdotus ia those of Tasmania. (It should be added that Australia possesses many winged pterostichines as well as these and other flightless genera.)
1961] Darlington — Australian Carabid Beetles 13
Of non-pterostichines, Pamborus is noteworthy. It is confined to eastern Australia and is one of the two known genera of the tribe Pamborini. (The other is monotypic Maoripamborus in New Zea- land— Brookes 1944.) The northern limit of Pamborus is probably near Cooktown. Four species of the genus occur in the main tropical rain forest system of North Queensland, chiefly or wholly in rain forest. Six other species occur in South Queensland and New South Wales. Some of them occur mainly in (sub) tropical rain forest, but viridis inhabits savannah woodland and some other species occur in open woods as well as rain forest, and some enter south temperate rain forest on the high plateaus of north-central New South Wales. The southern limit of the genus is near the Shoalhaven River about 70 miles south of Sydney. (Old records for Victoria are probably errors.)
The genus Mystropomus is the only Australian representative of the pantropical tribe Ozaenini. The genus is confined to eastern Australia. Its northern limit is between Daintree and Cooktown. A single species (two subspecies) occurs throughout the main tropical rain forest sys- tem of North Queensland, and is apparently confined to rain forest. Another, variable species (two subspecies) occurs in the subtropical rain forest system, and extends into more open woodland. The south- ern limit of the genus is apparently near Sydney.
These five genera dominate the flightless geophile carabid faunas of the main tropical and subtropical rain forest systems of eastern Australia. Their distribution is notable in several ways. All five genera reach an approximately common northern limit, north of Dain- tree and south of or near Cooktown. All five genera are widely dis- tributed both in the main tropical and in the subtropical rain forest systems. These two forest systems are separated by a wide barrier of comparatively dry, open forest in which is one important “island” of rain forest, on the Eungella Range west of Mackay, and all five of the genera in question are represented there.4 In the tropics, these genera occur only or chiefly in rain forest,5 although most of them enter opener forest too in the south temperate zone.
4Of the 5 genera in question on the Eungella Range, the one Pamborus has close relatives in both North and South Queensland. The one Mystropomus is a South Queensland species. Of 2 Trichosternus, one probably belongs to a South Queensland group and the other is doubtful. The one Notonomus be- longs to a North Queensland group. And the one Leiradira belongs to a South Queensland subgenus. These genera in the Eungella rain forest there- fore show 2 close ties with North Queensland (in Pamborus and Notonomus) and 4 with South Queensland.
5 Trichosternus cordatus Chd. occurs outside rain forest in the southern edge of the tropics.
14
Psyche
[March
Southward, through New South Wales, rain forest of (sub) tropical type diminishes in area and quality, and the Carabidae associated with it diminish too. Of the five genera just discussed, Leiradira may not extend south of the Dorrigo. T richosternus , Mystropomus , and Pam- borus go a little farther south, reaching different limits probably in this order, but do not reach Victoria. And N otonomus reaches Vic- toria (in numbers) and Tasmania (only one stock in rain forest). Toward their southern limits, all these genera, except Leiradira , occur not only in tropical-type rain forest but also in opener forest, and all, except again Leiradira , have entered or even evolved endemic species in south temperate rain forest on the Dorrigo-Ebor plateau and the Mt. Royal Range.
Transition in Australia: North from Tasmania
The ground-living Carabidae of the south temperate rain forest of Tasmania are dominated by or include flightless genera of four special tribes in addition to the more widely distributed Pterostichini, Licinini, etc.
The tribe Broscini is well represented in both the north and the south temperate zones of the world (Ball 1956) but is absent in the tropics or nearly so. Some northern broscines have well developed wings, but I think that all those of the southern hemisphere have atro- phied wings and are flightless. Four genera occur in Tasmania. Promecoderus is represented there by several rain forest species and by other species that live in drier, opener woodland. The genus is widely distributed across southern Australia, but chiefly in dry forest and arid country, although one or two species occur in rain forest in Victoria. Of the other Tasmanian genera, Chylnus is confined to Tasmania, in wet forest. Per cosoma occurs in Tasmania and the mountains of southeastern Victoria, in wet forest. And Eurylychnus occurs in Tasmania, southern Victoria etc. including the Otway Ranges, and east and north into southern New South Wales, and two separate stocks of the genus have species isolated (chiefly in south temperate rain forest) on the Mt. Royal Range and the Dorrigo-Ebor plateau. The latter is the northern limit of wet-forest broscines in Australia.
The tribe Trechini (subfamily Trechinae of Jeannel 1926-1928)
Explanation of Plate 4
Fig. 7. Diagram of transition of selected flightless geophile Carabidae in rain forests of eastern Australia. The 5 genera at bottom of the diagram are primarily tropical and subtropical ; the other genera and tribes, pri- marily south temperate. See text for further details.
Psyche, 1961
Vol. 68, Plate 4
Tasmania
i6
Psyche
[March
is world-wide in distribution. It includes both flying and flightless genera, but the flying genera usually live beside standing or running water and are not forest-living geophiles. Flightless geophile Trechini are numerous both north and south of the tropics. In Tasmania they are numerous in south temperate rain forest and hardly enter other habitats at low altitudes, although some occur in open country above timber line, on cold mountain tops. Flightless Trechini are less numer- ous but still widely scattered in wet forests and on mountain tops in southern Victoria, including the Otway Ranges (Moore i960), east nearly to the New South Wales border and north to Mt. Kosciusko in southern New South Wales, and endemic species perhaps represent- ing one original flightless stock of spotted " Trechus ” are isolated on the Mount Royal Range, the Dorrigo-Ebor plateau, and the Mc- Pherson Range on the Queensland border.6
The tribe Migadopini (Jeannel 1938; Darlington i960, p. 663) is confined to the southern hemisphere, with different genera localized in Tasmania and southeastern Australia, New Zealand, and the south- ern tip of South America, etc. Two flightless genera of the tribe occur in Tasmanian rain forest: Calyptogonia is confined to Tasmania; Stichonotus extends to the mainland, but only to the Otway Ranges. A third Australian genus of the tribe is known from a single specimen collected long ago near Kiama south of Sydney, and a fourth genus occurs still farther north, in subtropical forest on the low (c. 2,000 ft.) Comboyne plateau at about 3 1 0 35' S. This last genus, Decogmus, differs from all other Migadopini in being winged.
Finally, the flightless tribe Agonicini is confined to Tasmania and southeastern Australia (Moore i960). There are two genera. One is widely distributed in Tasmania and occurs also in the mountains of southern Victoria east of Melbourne (B. P. Moore, in letter). The other is confined to the mainland, including the Otway Ranges and the “Victorian Alps,” north to Mt. Kosciusko. Agonicines live on the ground in rain forest, and sometimes in open snow gum woods on mountains.
Although there are other Carabidae in Tasmanian rain forests (especially various Pterostichini and Licinini) the four tribes just discussed make up a large part, and zoogeographically the most im- portant part, of the flightless wet forest Tasmanian carabid fauna. It will be seen from details given above that all four tribes occur both in
6 A second “Trechus”, diemensis Bates, extends from Tasmania and south- eastern Australia north to the McPherson Range, but this species is winged or dimorphic.
1961]
Darlington — Australian Carabid Beetles
7
Tasmania and on the adjacent mainland of Australia, but that they all diminish rapidly northward.
The transition of selected elements of the flightless geophile carabid faunas of tropical and south temperate rain forests is diagrammed in Fig. 7-
Isolated Australian Faunules
To return to the five carabid genera discussed above as characteristic of the main tropical and subtropical rain forests, these genera have dis- tributions that are alike in many details. Within the main (base-of- peninsular) tropical rain forest system, they all have almost the same northern limits and (excepting perhaps Leiradira ) the same southern limits. All are represented on the Eungella Range. In South Queens- land, all apparently find their northern limit on Mt. Jacob (except that Trichosternus cordatus extends farther north in drier woodland), and all extend well into New South Wales, although they reach different limits there. They illustrate a general fact, that the carabid faunas of the main tropical and subtropical rain forest systems of east- ern Australia, although separated by several hundred miles of com- paratively dry country, are fundamentally similar, dominated by the same tribes, and share many genera some of which coincide remark- ably in details of distribution, although some other genera and most species are different. However three isolated pieces of Australian rain forest have carabid faunules that do not flt into this main pattern. They are the tip-of-peninsular and mid-peninsular rain forests of Cape York and the rain forest on the Elliot Range south of Towns- ville.
The tip-of-peninsular tract is light rain forest and is limited both botanically and zoologically. For example, st’nging trees (Laportea) , which occur in other Australian rain forests and in New Guinea, are apparently absent in the tip-of-peninsular forest, and land leeches and itch mites, which are pests in rain forest elsewhere, are apparently absent in the tip-of-peninsular tract. The winged Carabidae of this tract are not remarkable, except that they include New Guinean species. But the flightless Carabidae form a faunule wholly different from that of any other rain forest, consisting (so far as I could find) of only two flightless species. One is Mecynognathus dameli Mach, an enormous carabid, the largest males 2V2 inches long with mandibles like stag beetles. The genus occurs nowhere else on earth, although it may be rather closely related to Paranurus (see below). The other is a large flightless Clivina (probably kershawi SI.), which is fairly common both in the rain forest and in adjacent savannah woodland. The nature of this forest and of its flightless Carabidae suggests that
i8
Psyche
[March
the tip-of-peninsular tract is not a remnant of a larger, continuous rain forest but has been constituted or reconstituted separately, by gradual accumulation of a limited variety of plants and animals.
The mid-peninsular rain forests of Cape York are heavier and more extensive than the tip-of-peninsular tract, more like the base-of- peninsular forests at least superficially, but their flightless Carabidae form a second independent faunule. None of the flightless genera characteristic of the other rain forests is represented in the mid- peninsular system. In their place is a single large species of Parcimirus. This is a genus of probably only one, geographically variable species, which occurs from the tip of Cape York (and islands off the tip) south to below Cairns mainly in good savannah woodland. In most parts of its range it apparently does not enter rain forest, but it has done so in the mid-peninsular system, where it is now widely distributed. It seems to have invaded this system recently. An earlier invasion of the tip-of-peninsular rain forest by the ancestral stock of Paranurus may have produced Mecynognathus . There is also in the mid-peninsular rain forest a flightless Coptocarpus, but it is small and rare and I am not sure of its habitat or relationships. And also in this forest is a large form of Lesticus chloronotus Chd. It is winged, but its distri- bution and behavior suggest that it may eventually become flightless, as several stocks of the same genus have done in New Guinea. The Carabidae, then, suggest that the mid-peninsular rain forest has not been connected with the main base-of-peninsular system but, like the tip-of-peninsular tract, has derived or is deriving its flightless Carabi- dae independently.
The rain forest on the Elliot Range is poorly known. The only insect collecting ever done in it, so far as I know, was done March 2, 1958, when my son and I climbed from Double Creek to near the peak of Sharp Elliot and worked for three or four hours in the forest there. It seemed to be real but rather light rain forest. We found there series of two conspicuous flightless Carabidae: a very big Nurus and a Notonomus, both endemic. No trace of the four other genera (other than N otonomus) discussed above as characteristic of the main tropical and subtropical rain forests of Australia was found. Judging from my experience elsewhere, we would probably have found speci- mens or fragments of other species if the carabid fauna were diverse. I think, therefore, that the rain forest of the Elliot Range probably has a limited, endemic faunule of flightless Carabidae presumably received across a barrier and not by way of continuous rain forest. The valley that separates the Elliot Range from the main mountain
1961]
Darlington — Australian Carabid Beetles
19
system of North Queensland is not much more than ten miles wide, but it seems to have been a more effective barrier than the much wider gaps of dry hilly country between the North Queensland, Eungella, and South Queensland rain forest areas.
Summary of Transition from New Guinea to Tasmania
The transition of wet forest carabid faunas from New Guinea to Tasmania involves two main changes. First, between the rain forests of New Guinea and those of tropical Australia is a complete change of flightless stocks of Carabidae and also a change from Agonini to Pterostichini as dominant tribes, although the change is overlain and superficially concealed by many winged species and genera of Carabi- dae that are common to New Guinea and Australia and that form a broad and complex transition, not fully described here. Between the tropical rain forests of North Queensland and the subtropical ones of South Queensland etc. are very many changes of species and genera but no fundamental change in the nature of faunas or in dominant groups. The second main change is farther south, and is a complex transition from tropical to south temperate groups. The area of transi- tion (of overlapping and mixing of faunal elements) is from the southern edge of Queensland to Tasmania. And the transition in- volves not only changes of species and genera but a second partial change of dominant tribes, from Pterostichini as principal dominants to (in Tasmania) dominance shared by Broscini and Trechini (and Licinini) as well as some Pterostichini. This change has been de- scribed as it occurs among selected flightless geophile Carabidae, but it is reinforced and made more complex by changes of winged Carabi- dae too.
The whole transition of wet forest carabid faunas from New Guinea to Tasmania might be described as a very irregular stepcline of flightless groups overlain by a more regular transition (or cline of many smaller steps) of winged groups. The flightless Carabidae of the isolated rain forests of Cape York and the Elliot Range are outside the main pattern and complicate it, and of course the situa- tion as a whole is much more complex in detail than I can describe here.
Historical Duplications : Two Land Bridges
It is a good working principle of zoogeography that situations should be analyzed first by study of the best known and most significant groups of animals, especially mammals, but that other groups may add important details to what the mammals show. In the present case, two former land bridges are involved: from New Guinea to Australia
20
Psyche
[March
and from Australia to Tasmania. Mammals show, by occurrence of many identical or closely related species on opposite sides of the exist- ing water gaps, that both bridges did exist recently and that some forest-living animals crossed both of them. Carabidae show additional, different things about the two bridges. In the case of the Australian- Tasmanian bridge, the Carabidae agree with the mammals. Many wet forest Carabidae including many flightless ones evidently crossed this bridge without meeting important ecological barriers, although cold climate stopped some other animals, especially some reptiles (Darlington i960, p. 659). In the case of the New Guinea-Austral- ian bridge, however, the flightless rain forest Carabidae show that there was an ecological barrier upon the land, and that the barrier existed for a long time. New Guinea and Australia cannot have been connected by a continuously rain-forested ridge within the time of existing carabid faunas. The recent connection was evidently low and rain forest was probably not continuous across it, although it was nearly enough continuous to allow certain forest trees, mammals, birds, and winged insects to get across. These organisms probably crossed the bridge by way of more or less separate forest “stepping- stones” and strips of gallery forest that did not allow continuous passage of flightless rain forest Carabidae, which do not disperse easily across even narrow gaps of unsuitable land. Rain forest is discontinuous on Cape York now. The Carabidae suggest that it has been so for a long time in the past, and that conditions on Cape York now are like the conditions that existed on the land bridge when New Guinea and Australia were connected.
Historical Implications : Climatic Fluctuations The present distribution of wet forest Carabidae shows that many of them have been able to move up or down the eastern edge of Australia between North and South Queensland, across what are now wide gaps of comparatively dry country. The degree of relationships of different Carabidae in the tropical and subtropical rain forest systems varies. In some cases ( e . g. Pamborus of the tropicus group) the North and South Queensland representatives of single original stocks are only slightly differentiated, but in other cases ( e . g. Leira- dira and its allies) they have diverged as subgenera or genera. This suggests either several periods of dispersal and isolation, accompanying fluctuations of rainfall and rain forest, or occasional trickling of dominant wet forest Carabidae across the drier gaps of central Queens- land. In either case wet forest Carabidae seem to have followed a rather narrow path along the continental divide, and have usually
1961]
Darlington — Australian Carabid Beetles
21
not been able to reach such slightly isolated places as the rain forest on the Elliot Range. The whole pattern, of occasional or limited exchange between North and South Queensland and of isolation of endemic faunules on the Elliot Range and in the Cape York rain forests, is consistent with climates and forests fluctuating only within moderate limits, not profoundly.
Ecological Correlations
It is a fact not sufficiently understood by some zoogeographers that the climatic zones, the differences between tropical and cool temperate climates, are very important to Carabidae and other insects. In eastern Australia, where climate is the only permanent barrier to dis- persal, many old groups of Carabidae are confined to either the tropical (including subtropical) or the cooler south temperate areas. Evident- ly whole tribes may persist for long periods in small areas protected only by climatic barriers, and even dominant tribes do not always easily cross from one climatic zone to another,
Carabid distribution is correlated with climate and ecology in sev- eral more specific ways. For example some rain forest Carabidae, including five genera specially considered above, seem to be more strict- ly limited to rain forest in the tropics than in the south temperate zone. This suggests that ecological factors are more intense in the tropics, as they may well be if temperature and evaporation rates are involved. That ecological factors are intense in the tropics is suggested also by groups of Carabidae that occur in diverse habitats in the temperate zones but enter or cross the tropics only when associated with surface water, which probably tempers the intensity of tropical climate. I have discussed this elsewhere (1959, especially pp. 332, 342). In Australia, for example, the only Trechini that occur in the tropics are winged hydrophiles: Perileptus and Trechodes by running water and Trechobembix (which extends north to Cairns) in deep swamps. Mecyclothorax occurs in many habitats in temperate southern Aus- tralia, but I found only one species (apparently cordicollis SI.) in the tropics, in thick vegetation over deep, cool water on the Atherton Tableland. And Notagonum (“ Agonum ”) submetallicum (White), which, though always associated with water, occurs in a variety of waterside habitats in both humid and arid parts of south temperate Australia, I found in the tropics (Atherton Tableland) only in thick vegetation over cool, moving water.
There is also a notable correlation of wings and flight of Carabi- dae with climate and altitude. Carabidae (mostly geophiles) often become flightless at low altitudes in temperate climates, and on moun-
22
Psyche
[March
tains everywhere, but rarely at low altitudes in the tropics. The few- ness of flightless Carabidae at low altitudes in New Guinea is an example. I have discussed this subject, with other examples, else- where (1943).
Finally there is a partial correlation between size of Carabidae and climate. Very large Carabidae (over 1 to 2^/2 inches long) are numer- ous in the forests of warm temperate to tropical eastern Australia but relatively few or absent in both cool temperate Tasmania and wholly tropical New Guinea. If Catadromus tenebrioid.es (ol.) is introduced, as I think it is, the largest carabids in New Guinea are hardly an inch long and few are that large. I suspect that this correlation has a complex ecological basis which may include direct action of physiologi- cal factors, correlation of size with state of wings and flight, and competition with other insects. Of insects that might compete with carabids, ants are most obvious. I have suggested (1943, p. 42, Fig. 4) that ants may take the place of most flightless geophile Carabidae especially in the lowland tropics.
Geographical History of Carabidae
Carabidae, like other old, complex groups of animals (mammals etc.), have presumably had complex geographical histories, with suc- cessive dominant groups evolving, spreading over the world, and replacing older groups. The present distribution of Carabidae in the Australian Region may reflect this. Some localized tribes that are now confined to the cool south temperate zone may be remnants of an ancient fauna (see Darlington i960 for further discussion of some of these groups). Pterostichini, now dominant in most of Aus- tralia, may be more recent and may be replacing more ancient Carabi- dae. And Agonini may be still more recent, now dominant in New Guinea (and in the whole tropical Asiatic- Australian area), and spreading to Australia.
Pterostichini and Agonini tend, as dominant tribes, to be com- plementary over the world as a whole. I have discussed this before (1956, pp. 1-3), but what I said then is worth repeating briefly now, with counts of species brought up to date. Both tribes are cosmopoli- tan, but unevenly so. In some parts of the world they occur in nearly equal numbers, in others, one tribe or the other is overwhelmingly dominant. The tribes tend to be complementary within the Australian Region, as already indicated. In Australia itself (with Tasmania) Pterostichini are dominant, with more than 350 known species against probably less than 20 species of Agonini, a ratio of nearly 40/1. But
1961]
Darlington — Australian Carabid Beetles
23
in New Guinea Agonini are dominant, with considerably more than ioo known full species (some discovered since my 1952 paper) against about 40 species of Pterostichini (manuscript in press), a reversed ratio of about 3/1.
One reason for the number of Agonini in New Guinea is that species of this tribe have multiplied on the mountains there. In Australia, however, Pterostichini, not Agonini, have multiplied in what seem to be comparable habitats on the mountains. This difference can hardly be accounted for in simple ecological terms but is probably due to a complex combination of ecological, historical, and geographical factors. Over the world as a whole, there is a tendency for Agonini to be better represented in the tropics; Pterostichini, in the temperate zones. Also it is probable that Agonini, which are phylogenetically less diverse, are more recent in origin than Pterosti- chini and that they have dispersed more recently. It is therefore likely that Pterostichini are dominant in Australia partly because Australia is more temperate than tropical in climate and partly because Pterostichini reached Australia before Agonini did, and it is likely that Agonini are dominant in New Guinea partly because the climate there is fully tropical and partly because the carabid fauna of New Guinea is more recent in its origins than that of Australia, as I think it is. Add to this that the mountain carabid faunas of Australia and New Guinea have been derived independently, each from the lowland fauna adjacent to it, and not by dispersal along a connecting mountain chain, and we have an adequate and probably correct explanation of the great difference in composition of the carabid faunas on the mountains of Australia and New Guinea.
As to direction of recent movements of Carabidae, movements of (winged) species have evidently occurred in both directions between Australia and New Guinea, although I cannot take space to give details now. Movements have apparently occurred also in both direc- tions between the tropical and subtropical forests of Australia. This is indicated by the relationship of the species now on the Eungella Range (p. 13), although I am not ready to give further details now. South of the tropics, patterns of distribution (Fig. 7) suggest withdrawal of cool temperate groups and southward spreading of tropical or subtropical groups. This is probably primarily an adjust- ment to recent warming of climate rather than an invasion of south temperate habitats by tropical Carabidae, although Pamborus , Tri- chosternus , and Notonomus have invaded N othofagus forest on high plateaus in New South Wales.
24
Psyche
[March
References
Ball, G. E. 1956. ... on the classification of the tribe Broscini. . . Coleop-
terists’ Bull., 10: 33-52.
Brass, L. J. 1953. Results of the Archbold Expeditions. No. 68. Summary of the 19+8 Cape York (Australia) Expedition. Bull. American Mus. Nat. Hist., 102: 135-206.
Brookes, A. E. 1944. \J\iaoripamborus.~\ Trans. Proc. R. Soc. New Zealand, 73 : 262.
CSIRO (Commonwealth Scientific and Industrial Research Organization) 1950. The Australian environment (2nd ed.). Melbourne. 183 pp.
Darlington, P. J., Jr. 1943. Carabidae of mountains and islands. . . Ecological Monographs, 13: 37-61.
1952. The carabid beetles of New Guinea. Part 2. The Agonini. Bull. Mus. Comp. Zool., 107: 87-252.
1953. Australian carabid beetles II. Some new Pterostichini. Psyche, 60: 90-101.
1956. Australian carabid beetles III. Notes on the Agonini. Psyche, 63 : 1-10.
1959. The Bembidion and Trechus of the Malay Archipelago. Pacific Insects, 1: 331-345.
1960. The zoogeography of the southern cold tem- perate zone. Proc. R. Soc. (London) (B), 152: 659-668.
1961. Australian carabid beetles IV. List of locali- ties, 1956-1958. Psyche, 67: 111-126.
Jeannel, R. 1926-1930. Monographic des Trechinae, L’Abeille, Vol. 32, No. 3; Vol. 33; Vol. 35; supplement, Vol. 34, No. 2.
1938. Les migadopides . . . une lignee sub-antarctic. Rev. francaise d’Ent., 5: 1-55.
Moore, B. P. 1960. Studies on Australian Carabidae (Coleoptera) — 1. New species of the tribes Agonicini, Trechini, and Pterostichini. Proc. R. Ent. Soc. London (B), 29: 165-169.
Womersley, J. S., & J. B. McAdam. 1957. The forests and forest condi- tions in the Territories of Papua and New Guinea. British Commonwealth Forestry Conference in Australia, 1957. 62 pp.
A RECONSIDERATION OF THE GENUS EPIPOMPILUS (HYMENOPTERA: POMPILIDAE)1
By Howard E. Evans Museum of Comparative Zoology
The genus Epipompilus was described by Kohl in 1884, with maximiliani Kohl, from Mexico, as type. It was next treated by Ashmead in 1902, who at the same time described a related genus, Aulocostetkus with bifasciatus Ashmead, from “Peru”, as type, Haupt, in 1930, erected the tribe Epipompilini for these two genera and several others; the others were shortly thereafter removed to another tribe. In 1944 Bradley presented a revision of the American species of Epipompilus and Aulocostetkus. Ashmead, Haupt, and Bradley all separated the two genera by whether or not the eyes are hairy. Since Ashmead said that Epipompilus has glabrous eyes, it is clear that he was unfamiliar with the genus; and both Haupt and Bradley admit they had never seen the genus. Thus we have the curious phenomenon of a genus being treated by three persons, none of whom had ever seen any specimens belonging to the genus as he conceived it. As a matter of fact the eyes of maximiliani are hairy, and Epipompilus as conceived by these three workers is a nonexistent genus: in actuality the name Epipompilus is a senior synonym of Aulocostetkus.
This is only one of several sources of confusion in the genus. Ash- mead described Aulocostetkus by merely placing it in a key and listing bifasciatus n. sp. as type. His description of bifasciatus can be and has been considered valid, but he gives no information other than the generic characters and the type locality (“Peru”), not even as to color pattern, which is of much value in separating species in this genus. Haupt used Ashmead’s name for a specimen from Costa Rica, while Bradley, unable to find Ashmead’s type, followed Haupt while expressing doubt that he had correctly identified Ashmead’s species. However, there is a specimen in the U. S. National Museum labeled as Aulocostetkus bifasciatus Ashmead in Ashmead’s handwriting and marked as type of that species. But to add to the confusion this speci- men bears the locality Bahia, Brazil, not “Peru” as it should. Now Costa Rica (Haupt’s specimen) is actually closer to Peru than is Bahia, Brazil, but I find it hard to reason away the identification label in Ashmead’s handwriting. Specimens of this genus are so rare that one is unlikely to make an error in labeling; in fact I doubt if Ashmead
Published with the aid of a grant from the Museum of Comparative Zoology at Harvard College.
25
26
Psyche
[March
ever saw any other specimens of the genus. On the other hand, Ash- mead was a sufficiently careless person so that it is quite believable that he may have jotted down “Peru” when he meant “Brazil”. At any rate, I accept this as the type of hifasciatus and have presented a description of it below, along with a new name for Haupt’s specimen from Costa Rica.
Still further problems remain. Was Ashmead correct in placing maximiliani in the synonymy of Cresson’s aztecus, or was Bradley correct in resurrecting it? If Epipojnpilus and Aulocostethus are synonyms, what is the status of Banks’ Epicostethus, said to share some of the characters of both genera? What is the correct generic place- ment of Epipompilus msularis Kohl, from New Zealand? Finally, what is the male sex of Epipompilus ?
On the following pages I have presented a brief synopsis of Epipompilus in which answers to all of these questions are proposed. I do not mean to imply that all problems in the genus are solved : my synopsis is based on a mere 18 specimens of these exceedingly rare insects. There are doubtless undiscovered species, and the males of most of the species have yet to be found. But at least I hope that I have supplied a sounder framework for future studies than has previously been available.
Genus Epipompilus Kohl
Epipompilus Kohl, 1884. Verh. K. K. Zool.-Bot. Gesell. Wien, 34: 57. [Type species: Epipompilus maximiliani Kohl, 1884 ( — aztecus Cresson 1869) (designated by Ashmead, 1900)].
Aulocostethus Ashmead, 1902, Canad. Ent., 34: 132. [Type species: Auloco- stethus hifasciatus Ashmead, 1902 (monobasic and original designation)]. New synonymy.
Epicostethus Banks, 1947, Bull. Mus. Comp. Zool., 99: 445. [Type species: Epicostethus will'.amsi Banks, 1947 (monobasic)]. New synonymy.
Generic characters. — Maxillary palpi unusually elongate, antepe- nultimate segment the longest and distinctly longer than third antennal segment ; labial palpi with the penultimate segment broadly ovate, the ultimate segment attached to one side of it; mandibles stout, rather smooth, with a few setae but without a lamina on the inferior margin which subtends a fimbriate groove, inner margin with a single strong tooth well back from apex; labrum partially exserted, broad and short, apical margin of clypeus broadly truncate or arcuately concave ; flagellum with rather coarse, bristling, semi-erect pubescence, particu- larly on the inner side of the basal segments of the female and on the outer side of the entire flagellum of the male; head broader than high, front with distinct small punctures; eyes densely covered with short hairs (reduced and scarcely noticeable in some females and in
1961]
Evans — Genus Epipompilus
27
the known males) ; front rather swollen above, between the antennal bases abruptly declivous to the much lower plane of the area frontalis and clypeus. Pronotum short or fairly long, sides of disc rather prominent, streptaulus absent or ill-defined; mesoscutellum and met- anotum prominent medially, latter with distinct lateral foveae; postnotum of variable length, front and hind margins subparallel; propodeum with smooth contours, slope low and even, almost flat behind ; front femora of female slightly to quite noticeably incrassate ; front tibiae and tarsi without spines, middle and hind tibiae with or without scattered short spines; segments of front tarsus of female unusually short; claws slender, with a strong, subapical tooth which is nearly parallel to the apical tooth ; ultimate tarsal segments without spines beneath, pulvillar pads small but giving rise to some strong setulae. Hind wing with anal lobe small, about .3-. 5 as long as sub- median cell, anal vein extending very slightly or not at all beyond junction of transverse median vein, latter vein leaving it at an angle, oblique, meeting media much before origin of cubitus ; fore wing with venation extending relatively close to outer wing margin, marginal cell acute, removed from wing-tip much less than its own length; three submarginal cells present, second and third receiving recurrent veins near middle, third much wider at apex than at base. Abdomen fusiform, in the female somewhat depressed apically, apical sternite rather flat, even obscurely grooved medially; male with or without conspicuous brushes of hair on sternites four and five, subgenital plate of remarkable structure, forming a very slender, hairy process apically, its basal plate (morphological sternite 8) unusually broad; male genitalia with short, simple parameres, volsellae short-setose, not expanded apically, basal hooklets double, aedoeagus small and of simple structure.
Remarks. — Epipompilus possesses a remarkable array of unusual structural features ; if one follows the practice of Bradley and Arnold of splitting the Pompilinae into numerous tribes, there can be no question that the genus deserves a tribe of its own. Personally, I am much impressed with certain similarities with A porus and related genera: the pronotum is similar, the front legs of the female some- what incrassate, and the head shape and hairiness of the eyes suggestive of certain Aporini. The male genitalia suggests Allaporus , as does the venation of the hind wing. Any division of the Pompilinae into tribes can be no more than tentative until such time as the classification of the family from a world point of view is more satisfactorily worked out. In the meantime, I prefer to place Epipo?npilus in the Aporini.
As here construed, the genus Epipompilus is strictly Neotropical
28
Psyche
[March
is distribution. However, Kohl included a New Zealand species, insularis Kohl, in his conception of the genus, and various workers since have listed Epipompilus from the Australian region. I have studied two females determined by Banks as insularis and agreeing well with Kohl’s description of this species. There can be no question that this species is closely related to the several Neotropical species of Epipompilus. Indeed, it agrees well with the above diagnosis except in the following characters (the males are unknown) : maxillary palpi not greatly lengthened, about as usual in the family; labial palpi unmodified; eyes with only minute, scarcely noticeable hairs; trans- verse median vein of hind wing reaching media a short distance before origin of cubitus. The generic name Epipompiloides is here proposed for insularis Kohl, 1884. I know of no other species assignable to this genus, but the pompilid fauna of the Australian region is, of course, very inadequately known. This genus is related to Epipo?npi- lus and should be placed in the Aporini next to that genus.
Key to known species of the genus Epipompilus Males
Antennae moderately long, crenulate in profile; claws of front tarsus nearly alike; thorax in considerable part rufous; parameres of genitalia with extremely long hairs, abdominal venter with hair-
tufts 9. innub us n. sp.
Antennae very short, with coarse, dark pubescence but only very weakly crenulate in profile; outer claws of front tarsus much more strongly curved than inner claws; thorax black; parameres and abdominal venter with only short hairs 6. excelsus (Bradley)
Females
1. Wings wholly fuliginous; abdomen wholly rufous; pronotum
patterned with red and black (Florida and Bahamas)
I. pulcherrimus (Evans)
Wings hyaline, fore wing with two prominent brownish bands; abdomen not wholly rufous, more or less patterned with black,
rufous, and/or whitish, pronotum all black or all rufous .... 2
2. Abdomen black, with a pattern of whitish spots 3
Abdomen in part rufous, with or without whitish spots 5
3. Body wholly black except for a pair of whitish spots on second abdominal tergite; hind tibiae with a few short spines; prono- tum very short, subangulate behind (Ecuador)
2. williamsi ( Banks)
Thorax largely rufous ; hind tibiae without spines 4
1961]
Evans — Genus Epipompilus
29
4. Abdomen with whitish markings on tergites 2 and 5, none on
sternites; posterior lobes of pronotum rufous (Mexico)
3. aztecus ( Cresson )
Abdomen with whitish markings on tergites 2, 3, 4, and 5, also on sternites 2 and 3; posterior lobes of pronotum whitish (Brazil) 4. bifasciatus (Ashmead)
5. Size larger (8-1 1 mm.) ; abdomen with whitish markings on ter-
gite five, this tergite otherwise black (Brazil) 6
Size small (6-7 mm.) ; abdomen not marked with whitish on tergite five, tergites five and six brownish-ferruginous (Central America) 7
6. Antennae black; propodeum black except for limited yellowish
markings; hind tibiae unarmed 5* haupti (Aide)
Antennae rufous except darker apically; propodeum wholly rufo- castaneous; hind tibiae weakly spinose . 6. excelsus (Bradley)
7. Second abdominal segment with a pair of whitish spots; hind
tibiae unarmed (Panama) 7- delicatus Turner
Second abdominal tergite without whitish spots; hind tibiae with scattered, short spines (Costa Rica) 8. insolitus n. name
1. Epipompilus pulcherrimus (Evans) new combination Aulocostethus pulcherrimus Evans, 1955, Ent. News, 66: 150. [Type: $, Everglades Nat. Park, Florida, 30 December 1953 (U. S. Nat. Mus.)].
Remarks. — Since describing this species from the unique type, 1 have seen one additional specimen, from Mangrove Cay, Andros Is- land, May-June 1917 (W. M. Mann) [Amer. Mus. Nat. Hist.]. This specimen agrees well with the type except that it is smaller (about 5 mm. long, fore wing 4.3 mm.) and the pronotum has a broad median streak of pale rufous as well as being rufous anteriorly and posteriorly. This species is colored quite differently from any other. The pronotum is much shorter than in aztecus , nearly as short as in williamsi. The claws are the same as in aztecus and other species of the genus, my statement to the contrary in the original description being in error.
2. Epipompilus williamsi (Banks) new combination Epicostethus williamsi Banks, 1947, Bull. Mus. Comp. Zool., 99: 446. [Type: $. Banos, Oriente, Ecuador, 6000 feet, 30 Oct. (Mus. Comp. Zool.)].
Remarks. — Bank’s description is detailed and there is no reason to repeat it here. The pronotum is short and subangulate behind, the front femora are barely incrassate, and the hind tibiae have several spines. The first two of these characters are shared (more or less)
30
Psyche
[March
with pulcherrimus and delicatus, the third with insolitus , excelsus, and innubus. Thus the species is not as unique as Banks supposed, and his generic name must be added to the synonymy of Epipompilus. I have seen no specimens of this species other than Banks’ type.
3. Epipompilus aztecus (Cresson) new combination Ferreola azteca Cresson, 1869, Proc. Boston Soc. Nat. Hist., 12: 376 [Type: $, Veracruz, Mexico (Acad. Nat. Sci. Phila.)].
Epipompilus maximillian • Bradley, 1944, Trans. Amer. Ent. Soc., 70: 146 34: 57 [Type: 9, Cuernavaca, Mexico, 1871 (Bilimek) (Vienna Mus.)]. (Placed in synonymy with azteca by Ashmead, 1902). — Haupt, 1930, Mitt. Zool. Mus. Berlin, 16: 762.
Epipompilus maximilliani Bradley, 1944, Trans. Amer. Ent. Soc., 70: 146 (Misspelling of maximiliani Kohl).
Aulocostethus aztecus Bradley, 1944, ibid., p. 142.
Remarks. — Bradley has recently provided a detailed description of this species, drawn from Cresson’s type of azteca. He states that this specimen “does not at all agree with Kohl’s description of maxi- miliani\ Since I found myself unable to agree with this statement, I asked to borrow the type of maximiliani from the Vienna Museum for comparison. At first the type could not be located, but later Dr. R. M. Bohart visited the museum and at my request searched for and found it; Dr. Max Fischer then sent it to me by mail, and I took it to Philadelphia and compared it directly with the type of azteca. I am very much indebted to Drs. Bohart and Fischer for their assis- tance with this problem.
The two type specimens differ considerably in size, that of aztecus being much larger, 13.5 mm. long, fore wing 10 mm.; maximiliani measures 7.5 mm. long, fore wing 6 mm. The front femora are slight- ly more incrassate in aztecus (2.6 X as long as maximum width as compared to 2.75 X in maximiliani) . The whitish maculations are exactly the same in the two specimens, but they type of aztecus has the posterior third of the propodeum blackish and the middle and hind legs blackish except for the white spurs and white streaks on the hind tibiae; in the type of maximiliani the propodeum is wholly rufous, the middle and hind coxae rufous above, the middle and hind femora rufous, and the tibiae partially suffused with rufous. Besides these two specimens, I have seen one other, a female from Cuernavaca, taken by my wife inside the window of a house on March 24, 1959. This specimen is about the same size as the type of aztecus , from Veracruz, and the front femora are incrassate to the same degree. However, the propodeum is wholly rufous (as in the type of maximil- iani, also from Cuernavaca) and the leg coloration intermediate be- tween that of the two types (middle femora rufous, hind femora
1961]
Evans — Genus Epipompilus
3
black, middle tibiae slightly suffused with rufous but hind tibiae black and whitish, middle and hind coxae with a small amount of rufous above). There is no question at all in my mind that these three specimens are conspecific.
4. Epipompilus bifasciatus (Ashmead) new combination Aulocostethus bifasciatus Ashmead, 1902, Canad. Ent., 34: 132 [Type: $, Bahia, Brazil, 19 March 1883 (but stated by Ashmead to be “Peru”) (U. S. Nat. Mus., no. 58858)]. (Not Aulocostethus bifasciatus of Haupt, 1930; see no. 8. insolitus n. name).
Description of type female. — Length 11 mm., fore wing 8.7 mm. Head black ; inner orbits narrowly pale yellow up to emargination of eyes; antennae wholly brownish, darker apically; apical half of clypeus and labrum pale yellowish; mandibles dull rufous; palpi light reddish-brown. Thorax rufo-castaneous, except mesonotum with black streaks over wing bases, and the following pale yellow : posterior lobes of pronotum, extreme lower posterior corner of mesopleurum, posterior rim of propodeum (interrupted medially), and apical pos- terior parts of middle and hind coxae; legs otherwise reddish like thorax, middle and hind tibiae weakly infuscated, spurs all whitish. Abdomen black, spotted with pale yellow (almost white) as follows: two large lateral spots on tergite two, two much smaller spots on tergite three, two spots on tergite four slightly larger than those on three, two large spots on tergite five broadly connected by a basal band; also sternites two and three with small lateral spots. Wings bifasciate, hyaline with a strong band over the basal and transverse median veins and a broader band filling the marginal cell and extend- ing to the posterior wing margin. Body and legs clothed with short, white hair; eyes short-haired.
Clypeus broadly truncate; labrum small, exserted. First four antennal segments in a ratio of about 13:5:11:12, segment three .55 X upper interocular line. Head 1.2 X as broad as high; middle inter- ocular line .57 X width of head; upper interocular line .8 X lower interocular line. Ocelli in a broad, flat triangle, front angle greater than a right angle; postocellar line much greater than ocello-ocular line. Pronotum of moderate length, posterior margin subangulate. Propodeum with smooth contours except posterior slope finely trans- versely striolate and with rather long pubescence; median line not impressed. Posterior tibiae without spines. Fore wing with basal and transverse median veins interstitial; radial vein somewhat angulate at junction of second intercubital vein, marginal cell removed from wing- tip by about .7 its own length.
Remarks. — - This is a rather typical member of the genus, in fact
32
Psyche
[March
rather close to aztecus. As mentioned in the introduction, Ashmead provided no real description of the species, and Bradley did not see the type and therefore followed Haupt, who had misidentified the species. Presumably Ashmead was merely in error when he gave “Peru” as the type locality of the species, as the type is labeled in Ashmead’s handwriting.
5. Epipompilus haupti (Aide) new combination Aulocostethus haupti Arle, 1936, Festschr. fur Embrik Strand, 1: 514 [Type: 9, Serra do Realengo, Rio de Janeiro, Brazil, 7 Oct. 1934]. — Bradley, 1944, Trans. Amer. Ent. Soc., 70: 144. — Banks, 1947, Bull. Mus. Comp. Zool., 99: 445.
Remarks. — Bradley has provided a translation of the description of this species. I have studied the specimen mentioned by Banks (Campinas, Brazil). The legs of this specimen are more extensively rufous than described for the type, but there is agreement in most other details; the eyes of this species are more weakly hairy than is usual in the genus.
6. Epipojnpilus excelsus (Bradley) new combination Figs. 3 and 4
Aulocostethus excelsus Bradley, 1944, Trans. Amer. Ent. Soc., 70: 143 [Type: 9, Nova Teutonia, Santa Catarina, Brazil, 25 January 1939 (Mus. Comp. Zool.)]. — Wahis, 1957, Bull, Ann. Soc. R. Ent. Belg., 9 3: 47-49 (Remarks on color variation).
Remarks. — I have seen several additional females of this species from the type locality, as well as a female from Rio de Janeiro, October 1938 (R. C. Shannon) [U. S. Nat. Mus.]. The type, as well as all the other specimens I have seen, has a pair of connected whitish spots on the fifth tergite, Bradley’s description being in error on this point. Wahis has discussed this matter and also pointed out that some specimens have whitish markings on the second and sixth tergites.
A male Epipo?npilus in the Canadian National Collections, Ottawa, is almost certainly that of excelsus , even though it is colored very differently from the female. Like the type female excelms , it was taken at Nova Teutonia, Brazil, by Fritz Plaumann, in this case on 19 June 1946. The spinose hind tibiae, as well as the locality, suggest that this male belongs here rather than with haupti or bifas- ciatus.
Description of male . - — Length 5.8 mm., fore wing 4.8 mm. Body wholly shining black, with a weak bluish luster; face with a pair of small whitish spots beside and below the antennal sockets, next to the eyes; apical two thirds of mandibles whitish, the teeth rufous; palpi
1961]
Evans — Genus Epipo?npilus
33
light brown ; antennae dark brown ; tegulae dark brown ; front and middle legs with the coxae black, suffused with brown apically, re- mainder of these legs brown, middle femora with a tinge of rufous, front tibiae yellowish-brown ; hind legs wholly black except tibia with a sub-basal whitish spot which nearly encircles them ; fore wing weakly tinged with brownish, especially along the basal vein and on the apical third, setulae dark, veins and stigma brown.
Maxillary palpi with segments 3-6 in a ratio of about 15:19:15:13. Mandibles with a single large tooth well back from apex. Clypeus arcuately emarginate apically, exposing the small labrum. Eyes strongly convergent below, lower interocular line about .75 X upper interocular line; middle interocular line. .59 X width of head; head about 1. 1 5 X as wide as high; ocelli in a broad triangle, postocellar line 1.3 X ocello-ocular line. Front with distinct punctures which are separated by about their own diameters. Eyes with minute hairs except near the tops, where they are somewhat longer. First four antennal segments in a ratio of about 15:5:8:9, segment three about 1.6 X as long as thick; flagellum short, very weakly crenulate in profile, with coarse, dark pubescence which is especially long and suberect on the upper and outer sides of the basal segments.
Pronotum of moderate length, its posterior margin subangulate. Mesonotum wholly and uniformly covered with small punctures. Postnotum nearly as long as metanotum, polished, with a median impression and some weak basal striations. Propodeum with the slope low and even; median line not impressed. Femora slender; middle tibiae with a few spines, hind tibiae with many fairly strong spines above; all tarsi weakly spinose; longer spur of hind tibiae nearly as long as basitarsus. Claws with the inner tooth of all claws strong, sloping so that the claws appear bifid; outer claws of front tarsus much more strongly curved than inner claws. Fore wing with basal vein arising a very short distance beyond transverse median vein, basal part of basal vein distinctly arched; marginal cell large, acute apically, radial vein distinctly angulate at its junction with the second transverse cubital vein.
Abdomen fusiform, covered with short setae but without distinct ventral hair-brushes. Subgenital plate (fig. 3) consisting of a long, hairy apical process arising from a basal section which also bears some long hairs. Genitalia (fig. 4) with the parameres weakly setose; volsellae simple, weakly setose; basal hooklets large, double; para- penial lobes somewhat knobbed apically, very slightly exceeding the volsellae.
34
Psyche
[March
Fig. 1 Subgenital plate of Epipompilus innubus new species. Fig. 2 Gen- italia of E. innubus. Fig. 3 Subgenital plate of E. excelsus (Bradley). Fig. 4 Genitalia of E. excelsus. All figures show the ventral aspect.
1961]
Evans — Genus Epipompilus
35
7. Epipompilus delicatus Turner
Epipompilus delicatus Turner, 1917. Ann. Mag. Nat. Hist., (8) 20: 359 [Type:
$, Bugaba, Panama (Champion) (British Mus.)]. — Bradley, 1944, Trans.
Amer. Ent. Soc., 70: 146.
Remarks. — I have not seen the type of this species, nor had Brad- ley. It is a small species, comparable in size to pulcherrimus and insolitus. It is reported to have a short pronotum and unarmed hind tibiae, as well as a color pattern distinct from that of other species.
8. Epipompilus insolitus new name Aulocostcthus bifasciatus Haupt, 1930. Mitt. Zool. Mus. Berlin, 16: 763. —
Bradley, 1944, Trans. Amer. Ent. Soc., 70: 145 ( Nec Ashmead, 1902,
Canad. Ent., 34: 132; misidentification) .
Type. — <j>, Turrialba, Costa Rica (Coll. H. Haupt, Halle/Saale, Germany) .
Description (from Haupt, 1930).- — Length 7 mm. Yellowish- brown, the following black: head, apical half of antennae, tergites 1, 3, and 4, apex of hind femora, and outer base of hind tibia. Fore wing with two dark brown bands. Hind tibiae with a whitish area behind the black base, tips of front and hind coxae also whitish. Whole body with short, thick whitish hair ; eyes and wings hairy.
Wings (Haupt’s fig. 64) yellowish-hyaline, a brown transverse band before the middle and one before the apex. Fore wing with three cubital cells, the second somewhat pentagonal, the third trapezoidal, the latter removed from tip by its own length. Radial cell longer than second and third cubital cells together, also somewhat higher than these. Pterostigma cell-like, translucent, somewhat attenuate. Lower section of basal vein about twice as long as upper and weakly arched; transverse median vein interstitial. Hind wing with trans- verse median vein short, oblique, reaching media more than its own length before origin of cubitus.
Head thick, flattened and weakly concave immediately behind the eyes, temples barely developed. Ocelli large, forming a weakly acute angle in front, postocellar and ocello-ocular lines equal. Front strongly swollen, eyes thick, their inner margins subparallel, the eyes together about equal to width of front. Antennae filiform, relatively thick, third segment somewhat shorter than scape. Clypeus short, trapezoidal, its entire anterior margin weakly arcuately concave. Seg- ments of maxillary palpi very long, third segment about ten times as long as thick (Haupt’s fig. 65). Pronotum short, hind margin obtuse- ly angled, sides parallel, with distinct longitudinal swellings. Meso- scutum twice as long as pronotum medially; scutellum and metanotum
3 6
Psyche
[March
elevated. Postnotum distinct, half as long as metanotum, lightly impressed medially. Propodeum somewhat longer than broad, nar- rowed and less steeply sloping behind, without discernible sculpturing except indication of a median groove. Fore tarsi without a comb, second to fourth segments short, the second as long as broad, the following shorter. Hind tibiae with scattered, short spines. Claws slender, with a sharp tooth before the apex, also with a distinct fan of bristles (Haupt’s fig. 84 J) ; claw-comb with a very short plate, its barbules sparse, surpassing the pulvillus.
Re/narks . — I have not seen this species, but since it has been described and figured by Haupt in considerable detail, it seems de- sirable to provide a name for it.
9. Epipompilus innubus new species Figs. 1 and 2
Type. — cf , Cucharas, 750 m., Valley of Huallaga, Dpt. Huanuco, Peru, June 1954 (F. Woj'tkowski) [Coll. H. K. Townes].
Description. — Length 6 mm., fore wing 5.7 mm. Head black except as follows: inner orbits pale yellow up to middle of eyes; clypeus, labrum, and mandibles pale yellow, almost white ; palpi very light brown ; antennal sockets connected by a light yellow band ; first five antennal segments yellowish-brown below, dark brown above, rest of antenna nearly black. Thorax rufo-ferruginous except shining blue-black as follows: propleura and extreme anterior parts of prono- tum, mesosternum and anterior half of mesopleurum, sides of metano- tum, all of postnotum, all of metapleurum except upper anterior margin, all of propodeum except for sides of posterior rim, which are pale yellow. Coxae blackish except middle and hind coxae tipped with white; middle and hind trochanters blackish; front and middle legs otherwise light reddish-brown, hind legs nearly black except tarsi paler and tibiae with a white basal annulus; tibial spurs whitish except middle and hind spurs suffused with black basally. Abdomen shining blue-black except apical tergite ivory-white. Wings hyaline, with dark setulae, veins and stigma brown, fore wing weakly clouded in and about third submarginal cell.
Maxillary palpi very long, segments in a ratio of about 2:4:8:10:8:7. Mandibles rather smooth, with a few setae, inner margin with a strong tooth well back from apex. Labrum broad and short, truncate, exserted well beyond truncate apical margin of clypeus, latter about twice as broad as high. Front prominent above antennal orbits, narrow, middle interocular line .56 times width of head; head nearly 1.2 X as wide as high; ocelli in a broad, flat tri-
1961]
Evans — Genus Epipompilus
37
angle, postocellar line twice the ocello-ocular line. Front with distinct punctures which are separated by less than their own diameters. Eyes with very short, barely noticeable hairs. First four antennal segments in a ratio of about 15:5:13:12, segment three about twice as long as thick; each flagellar segment, but more particularly the middle ones, with a distinct swelling below and toward the base, giving the antennae a somewhat crenulate profile below.
Pronotum very short, its posterior margin subangulate. Mesonotum with distinct small punctures like the front ; postnotum smooth, trans- versely striate, about half as long as metanotum. Propodeum with even contours, median line weakly impressed, surface of declivity very finely transversely striolate. Femora not notably swollen; middle and hind tibiae with short spines scattered amongst the pubescence; longer spur of hind tibia nearly as long as basitarsus; claws of front and middle legs strongly dentate, those of hind legs obscurely dentate. Fore wing with basal vein arising well beyond junction of transverse median vein, basal part of basal vein strongly arched ; marginal cell large, acute, removed from wing-tip by only about half its length; other features of wing about as in other species of the genus.
Abdomen fusiform, covered with short setae ; sternites four and five each with a transverse brush of longer setae, longer on the sides than medially, the setae curved at their tips; genitalia also giving rise to some long setae which protrude from sides of subgenital plate. Sub- genital plate (fig. 1 ) of unusual form, consisting of a long, slender, hairy process arising from complex basal plates (the modified ultimate and penultimate sternites). Genitalia (fig. 2) with parameres short, bearing some very strong setae; volsellae weakly setose, narrow in ventral view but mesal surface wide and concave; basal hooklets double, unusually well separated ; parapenial lobes slightly shorter than volsellae; aedoeagus very small.
Remarks. — The spinose hind tibiae and short pronotum suggest williamsi as the possible female of this species, and the type localities of these two are not too far distant (Ecuador and Central Peru). However, the difference in coloration is great, and it seems to me best to consider the two distinct for the present.
THE REDISCOVERY AND PROBABLE PHYLOGENETIC POSITION OF PSILOPSOCUS (PSOCOPTERA)
By Edward L. Mockford Illinois State Normal University Normal, Illinois
The genus Psilopsocus has posed a puzzle to students of the Psocop- tera since the time of its discovery. The original description by Enderlein (1903:305) was based on a single specimen. Although adequate for identification, this description does not permit the genus to be placed beyond suborder in recent classifications. Enderlein placed Psilopsocus in the Mesopsocidae, but gave no reasons for this. Roesler (1944), apparently without re-examining the type, erected a new family for this genus and placed it in the group Epipsocetae on the basis of similarity of the lacinia in the Epipsocidae and Psilopso- cidae.
It is the purpose of this paper to describe a new species of Psilopsocus from the Philippine Islands, to add to the knowledge of the morph- ology of the genus, and to re-interpret its phylogenetic position.
Genus Psilopsocus Enderlein, 1903.
In addition to the characters included by Enderlein in the original description, the following characters are probably important in delimiting this genus:
1. Male phallic sclerotizations in the form of a simple frame with no indication of external parameres (fig. 2.).
2. Ovipositor valvulae complete, i.e. three pairs.
3. Female subgenital plate with a slender central process (fig. 1.).
4. Female paraproct with a decided elongation of the posterior margin (fig. 3.).
5. Male paraproct with a pointed apical process on its posterior margin (fig. 8.).
6. Labrum not of the Epipsocus type, lacking a pair of diagonal strap-like sclerites.
The character mentioned by Enderlein of lack of junction of the
Explanation of Plate 5
Psilopsocus nebulosus n. sp. Fig. 1, $, subgenital plate. Fig.2, $, hypandri- um and phallic frame. Fig. 3, $, left paraproct. Fig. 4, 9, sclerites of 9th abdominal sternite (dorsal view). Fig. 5, 9, ovipositor valvulae. Fig. 6, $, lacinial tip. Fig. 7, $, tarsal claw. Fig. 8, $, right paraproct. Fig. 9, $, tip of pedicel (Ped.) and base of first flagellar segment (fl).
38
Psyche, 1961
Vol. 68, Plate 5
Mockford — Psilopsocus
40
Psyche
[March
bases of Radius and Media + Cubitus in the hindwing is not found in the species described below, although a fold in the wing membrane running from the Radius just above this junction, to the wing base produces the appearance of two closely parallel veins.
In the species described below, the coxal organ is well developed, with both rasp and tympanum. The rasp is composed of tiny in- dentations with raised edges.
Psilopsocus nebulosus, new species Plate 5 and Text Fig. i
Diagnosis: differs from the generotype, P. nigricornis End., pos- sibly in number of teeth on lacinial tip (6 indicated for nigricornis , 1 1 in nebulosus) , number of ctenidia on basal segment of hind tarsus (15 in nigricornis , 16 to 19 in nebulosus) , shape of the pterostigma (less deep in nebulosus ), and details of the forewing markings (text figure 1), colorless areas being more numerous in nebulosus , but there being no colorless areas bordering the wing apex in this species.
Holotype $. Measurements (see table I.).
Morphology: IO/D (distance between eyes divided by greatest dorsal diameter of eye, method of Badonnel) = 0.78, PO (lateral diameter of eye from above divided by greatest dorsal diameter of eye) = 0.94. Apex of lacinia (fig. 6.) broad, with a distinct lateral and median cusp, the lateral broadest and bearing 10 denticles. Coeloconic sensilla of first flagellar segment two in number, both situated at the extreme base of the segment (fig. 9.). Distinct tarsal ctenidia present only on posterior tarsi, with a row of 17 on Ti and one each on T2 and T3 (Ti = basal tarsal segment). Hypandrium weakly sclerotized, its margin rounded except slightly flattened at the apex; this flattened apex slightly more heavily sclerotized than
Text figure 1. Psilopsocus nebulosus n. sp., $, photomicrograph of right forewing.
1961]
Mockford — Psilopsocus
4 I
the rest of the hypandrium and bearing a row of tiny cilia (fig. 2.). Phallic frame (fig. 2.) a simple ring, slightly oblong, rather uniform in width except somewhat thickened apically and for a short distance along each side. Within the frame are a pair of membranous lobes. Paraproct (fig. 8.) with the usual field of trichobothria extending obliquely downward from its antero-dorsal angle. Posterior margin of the paraproct extended as a lobe tipped with a pointed process.
Color (in alcohol) : compound eyes black. General body color dull ivory marked with medium to pale brown. Labrum, antennae, ocellar interval, and terminal two segments of maxillary palpi medium brown. Clypeal striations, cloudy spots around compound eyes and antennal bases, and a band bordering epicranial suture pale brown. Thorax irregularly mottled with large pale brown areas including most of the pleura and large areas of the tergal lobes. Brown areas of meso- thoracic tergal lobes darkest at their borders, pale in their centers. Legs dull ivory except medium brown on each coxa, distal end of each femur, distal end of each tibia, and all of each tarsus. Forewing hyaline, marked with extensive pale brown cloudy bands and spots as in text figure i. Abdomen ringed with irregular pale brown cloudy bands. Terminal segments medium brown.
Allotype ?. Measurements (see table I).
Morphology: IO/D = 1.56, PO = 0.72. Tarsal ctenidia present only on posterior tarsi with a row of 18 on Ti and one each on T2 and T3. Subgenital plate (fig. 1.) with basal pigmented area in the form of two widely diverging arms. Median process of subgenital plate broad basally, abruptly narrowing to a slender tongue about half-way toward its tip; the broad basal portion bearing two large setae and the slender apical portion bearing many setae of various sizes. Sclerites of the 9th abdominal sternite (fig. 4.) in the form
Table I. Length (in mm.) of various characters of Psilopsocus nebulosus
|
Character |
$ |
$ |
5 |
$ |
$ |
|
Entire body |
2.87 |
2.82 |
2.61 |
3.03 |
3-67 |
|
Forewing |
4.44 |
4.40 |
4.44 |
4.59 |
4.65 |
|
Hind tibia |
1.39 |
1.43 |
1.43 |
1.57 |
1.43 |
|
Hind tarsus, Ti* |
0.465 |
0.440 |
0.476 |
0.476 |
0.405 |
|
Hind tarsus, T2 |
O.071 |
0.071 |
0.059 |
0.071 |
0.059 |
|
Hind tarsus, T3 |
0.1 19 |
0.1 19 |
0.131 |
0.131 |
0.1 19 |
First posterior tarsal segment.
42
Psyche
[March
of a transverse ring with three straps radiating from it and a trans- verse strap basal to it. Ovipositor valvulae (fig. 5.) with first valvula slender; second valvula broad basally, terminating in a long, slender process; third valvula a broad flap bearing many setae. Paraproct (fig- 3-) with field of trichobothria in its antero-dorsal angle. Poste- rior margin of paraproct markedly protruding; ventral and posterior margins bearing many setae.
Variation: aside from variations noted in the descriptions and measurements (table I), four male paratypes have wing and body markings somewhat paler than the holotype male, but this may be due to tenerality. IO/D ratios for three male paratypes are 0.69,
0.70, and 0.70; PO ratios for these are 0.88, 0.85, and 0.9 1 ; numbers of ctenidia on posterior basal tarsal segment are 16, 19, and 18.
Nymph: one nymph taken with adults of P. nebulosus is with little doubt this species. The association is made on the basis of similarity in size, color, (except the distal two-thirds of the abdomen is dark brown in the nymph), and general body shape, also on the structure of the lacinia and tarsal claws. The lacinia is broad apically with several indistinct denticles on the outer cusp. The tarsal claw bears a preapical tooth and a pulvillus of medium width bent at a decided angle near its base, as in the adult (fig. 7.).
Type locality: Philippine Islands: Mindanao, east slope of Mt. McKinley, Davao Province, August and September, 1946, in mossy forest, elevation 6400 feet, H. Hoogstraal collector; holotype 3, allotype $ , 4 $ paratypes and one nymph, all in collection of Chicago Natural History Museum.
Discussion : Psilopsocus is apparently very close to the Myopsocidae.
The following points of similarity were noted:
1. Tarsal structure
a. Number of segments same.
b. Distribution of ctenidia same.
c. Both with preapical tooth on tarsal claw.
d. Pulvillus in both of medium width with a decided bend near its point of attachment.
2. Lacinial structure. The lacinial tip of Lichenomima sparsa has a broad lateral cusp bearing 13 stubby denticles, and a short median cusp, hence it shows marked similarity to the lacinial tip of Psilopsocus.
3. Male genitalia.
a. Hypandrial margin. The rounded hypandrial margin of Psilopsocus is similar to that of several species of Rhaptoneura , Phlotodes, and Lichenomima.
1961]
Mockford — Psilopsocus
43
b. Phallic frame. The simple, rounded phallic frame of Psilopsocus bearing a pair of lateral thickenings and enclosing a pair of membranous lobes is reminiscent of this structure in Lichenomima pauliani Bad. (Badonnel, 1955, fig. 529) and Rhaptoneura eatoni McL. (Badonnel, 1943, fig. 143) . It differs little from this structure in Lichenomima maxima Sm. (Smithers, 1957* fig- 6).
c. Paraproct. The paraproct of Psilopscus is similar to those of most Myopsocids in bearing a pointed process on its posterior margin. Although several species of Myopsocids have two such processes, there is only one in Phlotodes angolensis Bad. (Badonnel, 1955? fig- 514). The male paraproct of the latter species resembles that of Psilopsocus in several other respects, including shape of the field of trichobothria and presence of a roughened area antero-dorsad of this field.
4. Female genitalia.
a. Subgenital plate. The subgenital plate of Psilopsocus is sim- ilar in structure to that of Rhaptoneura. The resemblance is espe- cially marked in the case of R. africana Bad. (Badonnel, 1 955 > fig- 508). In both forms the pigmented basal area consists of a pair of widely diverging arms; the central process is broad basally, narrowing abruptly to a slender tongue which bears setae on or near its apex.
b. Ovipositor valvulae. Both Psilopsocus and the Myopsocids have three pairs of ovipositor valvulae. The second valvula terminates as a long, slender process in both groups, although this is generally much longer in the Myopsocidae than in Psilopsocus. The third valvula is a simple flap bearing many setae in both groups.
c. Sclerites of the ninth abdominal sternite. These sclerites in several Lichenomina species (Badonnel, 1955, figs. 522-525; Smithers, 1957, fig- 1 1 ) are composed of three main sclerotized areas, and are thus similar in appearance to the same group of sclerites in Psilopsocus nebulosus with their three radiating sclerotized straps.
d. Paraproct. The female paraproct is similar in shape, cilia- tion, and position of the field of trichobothria in the two groups.
The differences between the Myopsocidae and the Psilopsocidae are not great. The only ones which I have found are ( 1 ) presence in Myopsocidae of a connection between areola postica and medial stem in the forewing and absence of this in Psilopsocidae, (2) presence in Psilopsocidae of a spur vein from the pterostigma and its absence in Myopsocidae, and (3) much more complex markings of the forewing in Myopsocidae than in Psilopsocidae, with characteristic alternating dark and light areas on veins in the former group.
44
Psyche
[March
Pearman (1936) has designated the families Myopsocidae, Psocidae, and Thyrsophoridae as constituting the group Psocetae. To this group should be added the Psilopsocidae. It appears that this group repre- sents an ancient phylogenetic line within the suborder Psocomorpha, in which the most primitive forms share a broad, multidenticulate lacinial tip with the Group Epipsocetae. It seems likely that this type of lacinial tip, found also in the Amphientomidae (Suborder Trocto- morpha) was present in the earliest forms of the Suborder Psoco- morpha.
Acknowledgements
I wish to thank the officers of the Chicago Natural History Mu- seum, Mr. Henry Dybas in particular, for arranging the loan of material discussed in this paper. The accompanying photomicrograph (text figure 1) was made by Dr. Robert D. Weigel of the Depart- ment of Biological Sciences, Illinois State Normal University.
Literature Cited
Badonnel, A. 1943. Faune de France. Psocopteres. Paris, P. Lechevalier et Fils, 164 pp.
Badonnel, A. 1955. Psocopteres de l’Angola. Comp. Diamant. Angola Pub. Cult. 26: 1-267.
Enderlein, G. 1903. Die Copeognathen des Indo-australischen Faunenge- bietes. Ann. Mus. Hung. 1:179-344.
Pearman, J. V. 1936. The taxonomy of the Psocoptera: preliminary sketch. Proc. R. Ent. Soc. Lond. (B) 5: 58-62.
Roesler, R. 1944. Die Gattungen der Copeognathen. Stett. Ent. Zeit. 105: 117-166.
Smithers, C. N. 1957. Three new species of Myopsocidae (Psocoptera) from Natal. Proc. R. Ent. Soc. Lond. (B) 26:11-16.
CAMBRIDGE ENTOMOLOGICAL CLUB
A regular meeting of the Club is held on the second Tuesday of each month October through May at 3:00 p. m. in’ Room B-455, Biological Laboratories, Divinity Ave., Cambridge. Entomologists visiting the vicinity are cordially invited to attend.
The illustration on the front cover of this issue of Psyche is a drawing of a myrmecophilous weevil, Liometophilus manni Fall. The original figure was included in H. C. Fall’s article in Psyche, Volume 19, 1912, plate 3. The weevil itself was collected by the late W. M. Mann in the Huachuca Mts., Arizona.
BACK VOLUMES OF PSYCHE
The Cambridge Entomological Club is able to offer for sale the following volumes of Psyche.
Volumes 3, 4, 5, 6, 7, 8, each covering a period of three years. $8.00 each.
Volumes 10, 14, 17 to 26, each covering a single year, $2.00 each. Volumes 27 to 53, each covering a single year, $2.50.
Volumes 54 to 65, each covering a single year, $3.00.
Volumes 66 and 67, each covering a single year, $5.00.
Some other volumes, lacking certain issues, are also available (information upon request).
Orders for 10 or more volumes subject to a discount of 10%. All orders should be addressed to
F. M. Carpenter, Editor of Psyche, Biological Laboratories, Harvard University, Cambridge,, Mass.
FOR SALE
Classification of Insects, by C. T. Brues, A. L. Melander and F. M. Carpenter. Published in March. 1954, as volume 108 of the Bulletin of the Museum of Comparative Zoology, with 917 pages and 1219 figures. It consists of keys to the living and extinct families of insects, and to the living families of other terrestrial arthropods; and includes 270 pages of bibliographic references and an index of 76 pages. Price $9.00 (cloth bound and postpaid). Send orders to Museum of Comparative Zoology, Harvard College, Cambridge 38, Mass.
A JOURNAL OF ENTOMOLOGY
Established in 1874
Vol. 68 June-September, 1961 Nos. 2-3
CONTENTS
A New Earwig in the Genus Vostox (Dermaptera: Labiidae) from the Southwestern United States and Mexico. W. L. Nutting and A. B. Gurney 45
Some Comments on Walckenaer’s Names of American Spiders, Based on Abbot’s Drawings. H.fV. Levi and L. R. Levi 53
The Neotropical Species of the Ant Genus Strumigenys Fr. Smith: Miscel- laneous Concluding Studies. W . L. Brown , Jr. 58
A Note on the An tAAnamptogenys hartmani Wheeler. W . L. Brown, Jr. 69
Anthicus tobias Marseul, Another Tramp Species (Coleoptera: Anthi- cidae). F. G. Werner 70
Chemical and Biological Characterization of Venom of the Ant Solenopsis xyloni McCook. M. S. Blum , J. E. Roberts, Jr., and A. F. Novak 73
Mass Insect Control Programs: Four Case Histories. W. L. Brown, Jr 75
CAMBRIDGE ENTOMOLOGICAL CLUB
Officers for 1961-62
President ....J. J. T. Evans, Harvard University
Vice-President C. Walcott, Harvard University
Secretary A. R. Brady, Harvard University
Treasurer F. M. Carpenter, Harvard University
Executive Committee R. W. Taylor, Harvard University
S. K. Harris, Boston University
EDITORIAL BOARD OF PSYCHE F. M. Carpenter (Editor), Professor of Entomology , Harvard University
P. J. Darlington, Jr., Curator of Recent Insects, Museum of Com- parative Zoology
W. L. Brown, Jr., Assistant Professor of Entomology, Cornell University ; Associate in Entomology, Museum of Comparative Zoology
E. 0. Wilson, Associate Professor of Zoology , Harvard University H. W. Levi, Associate Curator of Arachnology, Museum of Com- parative Zoology
H. E. Evans, Associate Curator of Insects, Museum of Comparative Zoology
PSYCHE is published quarterly by the Cambridge Entomological Club, the issues appearing in March, June, September and December. Subscription price, per year, payable in advance: $4.50 to Club members, $5.00 to all other subscribers. Single copies, $1.25,
Checks and remittances should be addressed to Treasurer, Cambridge Ento- mological Club, 16 Divinity Avenue, Cambridge, Mass.
Orders for back volumes, missing numbers, notices of change of address, etc., should be sent to the Editorial Office of Psyche, Biological Laboratories, Har- vard University, Cambridge, Mass.
IMPORTANT NOTICE TO CONTRIBUTORS
Manuscripts intended for publication should be addressed to Professor F. M. Carpenter, Biological Laboratories, Harvard University, Cambridge, Mass.
Authors contributing articles over 8 printed pages in length may be required to bear a part of the extra expense, for additional pages. This expense will be that of typesetting only, which is about $6.00 per page. The actual cost of preparing cuts for all illustrations must be borne by contributors; the cost for full page plates from line drawings is ordinarily $12.00 each, and the full page half-tones, $18.00 each; smaller sizes in proportion.
AUTHOR’S SEPARATES
Reprints of articles may be secured by authors, if they are ordered at the: time proofs are received for corrections. A statement of their cost will be furnished by the Editor on application.
The March 1961 Psyche (Vol. 68, no. 1) was mailed August 2, 1961. The present issue of Psyche (Volume 68, nos. 2-3) is published with the aid of a National Science Foundation Grant (G 15817) to the Cambridge Entomological Club.
The Lexington Press. Inc., Lexington, Massachusetts
PSYCHE
Vol. 68 June-September, 1961 Nos. 2-3
A NEW EARWIG IN THE GENUS VOSTOX (DERMAPTERA: LABIIDAE)
FROM THE SOUTHWESTERN UNITED STATES AND MEXICO1
By W. L. Nutting2 and Ashley B. Gurney3
During the summer of 1958 a single male earwig was taken from a light trap in southwestern New Mexico and sent to the U. S. National Museum for identification. Apparently a new species of Vostox, it was put aside with the hope that more specimens might be collected. In the fall of 1959, during a study of the Dermaptera in the Univer- sity of Arizona collection, six adults and three nymphs of this same earwig were discovered among some undetermined specimens. A fur- ther search finally resulted in the completion of the series of six males, seven females, and three nymphs upon which the following description is based. This new earwig brings the total number of Dermaptera in the United States, both native and adventive, to 19 species and 1 subspecies.4 Probably not more than six or seven of them are repre- sentatives of our endemic fauna.
There are about seven previously described species of Vostox , of which only brunneipennis (Serville) occurs in the United States; the others are all Neotropical. V. brunneipennis ranges from Virginia. Indiana, and Illinois south to Florida and westward to eastern Texas, with a few records from Panama and the states of Vera Cruz and Sinaloa in Mexico. So far as the available material demonstrates, the new species ranges from southern New Mexico and Arizona into the Mexican states of Sinaloa and Baja California. Neither of the species
Arizona Agricultural Experiment Station Technical Paper No. 642.
department of Entomology, University of Arizona, Tucson.
3Entomology Research Division, Agricultural Research Service, U. S. De- partment of Agriculture, Washington, D. C.
“Several changes, which have occurred in the list of United States Dermap- tera since the paper by Gurney (1950) appeared, may be noted. Prolabia has been found to be a synonym of Marava, and M. wallacei (Dohrn) a synonym of M. arachidis (Yersin), the latter current combination replacing Prolabia arachidis (see Hincks, 1954). Pyragropsis buscki (Caudell), a recent addition to the list, occurs in Florida (Gurney, 1959). Prolabia pulchella (Serville) has been transferred to Laprobia, a genus described as new by Hincks (1960).
45
46
Psyche
[June-September
of V ostox found in the United States is apparently very common over most of its range, though fairly numerous specimens of hrunneipennis have been seen from Florida and other southeastern states, where it occurs beneath loose bark of trees. Spongovostox apicedentatus (Cau- dell) is similar superficially to the new species and is one of the commonest earwigs native to the southwestern United States and northwestern Mexico. The following keys serve to distinguish the latter and the two United States species of V ostox, in spite of their general similarity in habitus, size, and coloration.
Keys to Species of V ostox and Spongovostox Found in Continental United States (Males)
1. Forceps armed with a conspicuous subapical tooth
Spongovostox apicedentatus (Caudell)
Forceps armed with at least one conspicuous tooth at or consider- ably anterior to middle (if tooth is absent, forceps are definitely concave internally on basal third) 2
2. Forceps sparsely tuberculate beneath, not concave internally,
typically bearing a prominent, rounded tooth considerably an- terior to middle (if two prominent teeth occur, the smaller, secondary tooth is at the middle) ; pygidium as in Figs, io or ii ; parameres with conspicuous preapical curvature, Fig. 9.
V ostox hrunneipennis (Serville)
Forceps smooth beneath, generally conspicuously concave internal- ly on basal third, larger specimens with tooth near middle; pygidium as in Figs. 2, 3 or 7 ; parameres less conspicuously curved, Fig. 8. V ostox excavatus , new species
(Females)
1. Forceps armed with a basal, quadrate tooth, projecting but little
beyond dorso-internal margin; abdominal sterna moderately clothed with fine yellow-brown setae and bearing many long, brown setae on posterior margins (males and nymphs as well) ; suggestions of lateral folds on segments four, five, and some- times six (sometimes subtle but, when prominent, each fold bearing a long, light brown seta) ; pygidium much like Fig. 5.
Spongovostox apicedentatus (Caudell)
Forceps armed with a large, basal, quadrate tooth, projecting well beyond dorso-internal margin (Fig. 1) 2
2. Dorsal surface of anal segment with a scattering of prominent
tubercles over posterior third (Fig. 13) ; ventro-internal margin of forceps prominent and crenulate, dorso-internal margin
1961]
Nutting and Gurney — Genus Vostox
47
broadly rounded and beset with a few widely spaced tubercles, inner face thus scarcely concave for more than half its length ; pygidium as in Fig. 12, but scarcely diagnostic.
Vostox bru n n eipen n is (Serville)
Dorsal surface of anal segment comparatively smooth ; dorso- and ventro-internal margins of forceps prominent and closely set with small tubercles (almost crenulate), inner face thus dis- tinctly concave as a longitudinal groove nearly to tip ; pygidium as in Fig. 4. Vostox excavatus , new species
Vostox excavatus, new species Figures 1-8
Description. Male (holotype) : Size medium, form usual for genus; body depressed with sides of abdomen (except for slightly narrower segments 1 and 10) subparallel and as wide as elytra; abdomen minutely punctulate above and below, less so on segments 1-3, in- creasingly so posteriorly, body practically smooth elsewhere; fine, short setae rather densely covering labrum, antennae and limbs, but sparse on remainder of body including forceps, elytra and wing scales ; a few longer setae on posterior margin of head, anterior margin of pronotum, cephalic faces of femora, near bases of coxae, and on the posterior margins of all abdominal sterna except the last.
Head cordate in dorsal outline, with greatest width through the eyes equal to the median length ; occipital margin broadly and obtusely emarginate ; caudal angle of genae broadly rounded ; eyes not especially prominent, slightly shorter in length than the postocular portions of genae ; eedysial cleavage lines very faintly impressed ; antennae broken, one with 1 1, the other with 12 segments, the first segment equal to the sixth in length, considerably shorter than the fourth and fifth together.
Pronotum subquadrate, with greatest width at caudal third nearly equal to its median length, cephalic margin produced mesad to form a narrow cervical flange, laterocephalic angles obtuse and narrowly rounded, lateral margins straight and diverging slightly to the broadly rounded caudal margin, anterior two-thirds of disc convex with lateral margins flaring upward, thus forming shallow furrows which broaden and become confluent with the flattened posterior third of disc; median longitudinal sulcus moderately impressed on convex portion of disc, but becoming obsolete in posterior third.
Elytra with median length 2.1 times the greatest width of a single elytron ; lateral margins nearly straight and subparallel, humeral angles broadly rounded, distal margin subtruncate. Exposed portions
Psyche
[June-September
of wings projecting posteriorly almost one-half the median elytral length with external margins converging gradually to the truncated distal extremities.
Abdomen broadened slightly in the middle, with basal segment notably narrower than anal segment ; lateral folds moderately promi- nent on second and third terga ; posterior margin of terga four, five, and six bordered with small tubercles, becoming obsolete laterad ; anal segment transversely rectangulate with sides subparallel, posterior margin truncated, but with a small lobe laterad above dorso-internal margin of each forceps.
Forceps, as in figure 7, about three-fourths as long as the normally exposed portion of abdomen, relatively smooth, slender, and straight except for gentle incurving of the apical third, a prominent, slightly rounded tooth just anterior to middle on dorso-internal margin; inner faces rather strongly excavate anteriorly, this becoming obsolete proximad from tooth; ventro-internal margin bearing a few irregular- ly spaced tubercles anterior to tooth ; pygidium, as in figure 7, with sides parallel at base, converging acutely to the narrowly rounded apex; subgenital plate slightly less exposed than the last tergum, its lateral margins oblique and broadly rounded into the somewhat con- cave distal margin; concealed genitalia as in figure 8.
Femora moderately inflated, anterior pair most strongly so, and subequal in length to anterior tibiae; tarsi long, slender, their ventral margins (particularly of metatarsi) bearing numerous, stiff setae; posterior metatarsus subequal to the combined length of the remaining two tarsal segments, the ventral surface with 2 rows of setae along the outer (lateral) margin, inner (mesal) margin with 2 longitudinal rows and numerous shorter marginal setae which are arranged in about 12 to 15 short, oblique, comblike rows to give a “stepped” or “staircase” effect. (The combs are best seen on clean specimens, in a mesal view, with magnification of 50 or more times, in a strong light.)
Coloration: Similar to brunneipennis ; head, pronotum, median
third of wing scales and abdomen dark chestnut brown, paler on antennae, elytra, anal segment and forceps; outer two-thirds of wing scales yellowish-white; limbs honey yellow; eyes black.
Measurements (in millimeters) : Body length (exclusive of forceps and pygidium), 9; median length of head, 1.5; length of prono- tum, 1.5; median length of elytron, 2.5; internal length of exposed wing scale, 1.1; length of forceps, 4.3.
Female (allotype) : General form as in male, but somewhat more robust and differing as follows: head broader and longer; eyes larger
1961]
Nutting and Gurney — Genus Vostox
49
and slightly longer than cheeks; antennae broken, one with 9, the other with 12 segments; abdomen notably wider, but with anal segment considerably narrower than the basal segment; marginal tubercles absent from terga four, five, and six ; forceps typically shorter, stouter, and shaped as in figure 1, with a large quadrate tooth on dorso-internal margin at base, both dorso- and ventro-internal margins prominent and irregularly but closely set with small tubercles, inner faces thus distinctly concave nearly to tips ; pygidium shaped as in figure 4.
Coloration: Differs from male in no important respect except for being a shade darker over-all, especially on the anal segment and forceps.
Measurements (in millimeters) : Body length (exclusive of for- and pvgidium.), 9; median length of head, 1.5; length of pronotum, 1.5; median length of elytron, 2.5; internal length of exposed wing scale, 1.1 ; length of forceps, 4.3.
Variation: There are five male paratypes, three of which do not vary significantly in size from the type; the length (in mm.) of various parts of the smallest specimen (Tucson) follow: body 7.2, head 1.5, pronotum 1.2, elytron 2.1, wing scale 1.1, forceps 2.8. The eyes of all but the smallest agree with the type in being shorter than the genae, whereas in the smallest spec men they are slightly longer. The complete number of antennal segments varies from 12 to 16. The Tucson specimen also lacks the marginal tubercles on terga four, five, and six. Although the forceps of all are distinctively excavated, the large tooth is absent in the two smaller specimens( Fig. 6), and its position marked only by a tubercle in the third. The shape of the pygidium apparently varies considerably as in brunneipennis ; in two specimens it is unlike the type in that it is truncated at the tip (Figs. 2 and 3 ) . All genitalia are preserved in glycerol and show close agree- ment with those of the type in the shape of the parameres, details of the sclerotizcd armature of the basal vesicle, and the bend of the ejaculatory duct.
The six female paratypes show considerably less variation in size and configuration of characters; the lengths (in mm.) of various parts of the smallest specimen (“Vcnodio”) follow: body 8.6, head 1.6, pronotum 1.4, elytron 2.3, wing scale 1.2, forceps 2.2. The length of the eye of two agrees with the allotype in being longer than the cheeks, whereas in three it is shorter, and in the remaining specimen these measurements are equal. The number of antennal segments ranges from 13 to 16. Most of the paratypes vary but little in the shades of brown described above; however, the two females from
1961
Vol. 68, Plate 6
Nutting and Gurney — Vostox
V
1961]
Nutting and Gurney - — Genus Vostox
51
Baja California are a dark, smoky brown over-all, nearly black on the head, and diminishing posteriorly to a dark chestnut brown on the forceps. The appendages and outer portions of the wing scales are a lighter smoky brown.
Nymphs: Three nymphs, presumably collected with one of the adult males, are included with the paratypes. These specimens are probably more than half-grown, for they range in length from 6.5 to 7 mm., and each bears moderately developed, fused wing pads. The antennae are 10- to 12-segmented. Each of the first six abdominal sterna bears two long setae, which are conspicuously arranged in con- tralateral rows, one-third of the width of the abdomen from each margin. (This pattern may exist in the adult stage but is not evident in any of our specimens.) The smooth forceps range from 1.8 to 2.2 mm. in length but show none of the specializations of either sex beyond the minute tubercles along the dorso- and ventro-internal margins. Figure 5 shows the configuration of the pygidium which suggests that all three may be females. Their coloration is similar to the holotype, except that the outer two-thirds of both pairs of wing pads are dark brown and the inner third is a lighter, yellow-brown. Holotype: U.S.N.M. No. 65696
Type locality: Santa Catalina Mts. (2000-3000 ft.), Pima Co., Ariz.
The holotype male was collected by Andrew A. Nichol on Au- gust 15, 1924. In reply to a recent inquiry as to the exact locality, Dr. Nichol has recalled that it was in the lower parts of either Sabino Canyon (south slope of the range) or Canada del Oro (north and west slopes), probably the former. The allotype (U.S.N.M.) was taken under lights on the bridge over the Salt River (dry), Tempe, Maricopa Co., Ariz., on July 18, 1947, by Floyd G. Werner.
Paratypes: U. S. National Museum (1 cT, 1$, 2 nymphs) ; Depart- ment of Entomology, College of Agriculture, University of Arizona,
Explanation of Plate 6.
Figs. 1-8, Vostox excavatus , new species. 1. Forceps of female allotype, dorsal view. 2. Male pygidium (Tucson), dorsal view. 3. Male pygidium (Virden), dorsal view. 4. Pygidium of female allotype, ventral view. 5. Nymphal pygidium, ventral view. 6. Male forceps (Tucson), dorsal view. 7. Forceps and pygidium of male holotype, dorsal view. 8. Concealed genitalia of male holotype: Pm, paramere; Pn, penis; BV, basal vesicle; EjD, ejacula- tory duct. Figs. 9-14, Vostox brunneipennis (Serv.). 9. Left penis and para- mere (Gainesville, Fla.), dorsal view. 10. Male pygidium (Dallas, Tex.), dorsal view. 11. Male pygidium (Gainesville), dorsal view. 12. Female pygidium (Paris, Tex.), ventral view. 13. Female forceps (Mobile, Ala.), dorsal view. 14. Male forceps (Gainesville), dorsal view. Figs. 1-7, 10-14, xl5; Figs. 8 and 9, x34. (Drawings by senior author).
52
Psyche
[June-September
I ucson (id", 29, i nymph); Arizona State University, Tempe ( 1 cT ) ; Museum of Comparative Zoology, Cambridge. Mass. ( 1 d1 , i?) ; California Academy of Sciences, San Francisco, Calif, (id); Academy of Natural Sciences of Philadelphia (1$) ; British Museum (N. H.) , London, England ( i?) .
1 he paratypes are from the following localities :
NEW MEXICO: Virden, Hidalgo Co., one male, Aug. 27, 1958. light trap, G. L. Nielsen.
ARIZONA: Tempe, one male, Nov. 22, 1955, Jones; Tucson, Pima Co., one male and three nymphs, Dec. 4, 1924, C. T. Vorhies; one female, Nov. 7, 1939, Wayne Enloe; two females, Dec. 29, 1939, Tom Embleton ; Sabino Canyon, Sta. Catalina Mts., Pima Co., one male, July 25, 1955, at light, G. D. Butler and F. G. Werner; 2 mi. sw. Patagonia (4050 ft., Sonoita Creek bottom, willow-cottonwood), Sta. Cruz Co., one male, Aug. 21, 1949, F. H. Parker.
MEXICO: SINALOA, “Venodio”, one female, 1918, Kusche; BAJA CALIFORNIA, 25 mi. w. La Paz (ca. 500-foot plateau, relatively rich shrubby vegetation), one female, light trap, Aug. 30, 1959, K. W. Radford and F. G. Werner; 10 mi. sw. San Jose del Cabo (100 yd. from ocean in sandy wash, sparse shrubs), one female, light trap, Sept. 1, 1959, K. W. Radford and F. G. Werner. Aside from the few notes appended to the above localities, there is no information of any sort available on this apparently rare earwig. Morgan Hebard (1923, and other papers) described many Orthop- tera collected by J. A. Kusche in Sinaloa, at “Venvidio”, which probably is our “Venodio. ” Workers have been unable to locate either locality since, and Irving J. Cantrall, of the University of Michigan, has written us of having prepared a manuscript dealing with the Kusche locality. Thanks to Dr. Cantrall’s cooperation, we are able to report that Venadillo apparently is the correct name. This small town is 5 miles northeast of Mazatlan on Mexican Highway 15, which goes to Culiacan.
Literature Cited
Gurney, A. B.
1950. An African earwig new to the United States, and a corrected list of the Nearctic Dermaptera. Proc. Ent. Soc. Washington 52: 200-203.
1959. New records of Orthoptera and Dermaptera from the United States. Fla. Ent. 42: 75-80.
Hebard, M.
1922. Dermaptera and Orthoptera from the State of Sinaloa, Mexico, Part I. Trans. Amer. Ent. Soc. 48: 157-196.
Hincks, W. D.
1954. Notes on Dermaptera, I. Proc. R. Ent. Soc. Lond. (B), 23: 159-163.
1960. Notes on Dermaptera, IV. Ibid. 29: 155-159.
SOME COMMENTS ON WALCKENAER’S NAMES OF AMERICAN SPIDERS,
BASED ON ABBOT’S DRAWINGS1
By Herbert W. Levi and Lorna R. Levi Museum of Comparative Zoology, Harvard University
In 1887 McCook rediscovered the Abbot drawings, basis of many of Walckenaer’s spider descriptions, and initiated a controversy in spider nomenclature by synonymizing spider names then in use. Emerton replied that the drawings represent the spiders in so gen- eral and indefinite a way that identification would only increase the uncertainty of nomenclature. Banks’ comments about the Walck- enaer descriptions were blunt: “They rank with ‘hearsay evidence.’ I shall not use them nor list them; I shall ignore them.” Later Gertsch (1933) expressed the fear that these names would be re- vived and cause permanent instability: “The problem at hand is not the question of validity, which should be unchallenged, but one of recognition.” In 1944 Chamberlin and Ivie made a serious attempt to establish the Walckenaer names en masse. Their synonymies were accepted by Archer (1946, 1950), Levi (1954), and Levi and Field (1954), but not by Gertsch (1953). We were at first inclined to follow Chamberlin and Ivie in using the Walckenaer names, but during the course of the theridiid studies, had an opportunity to examine the Abbot drawings. We are convinced that in the Theri- diidae at least, Chamberlin and Ivie were ill-advised to attempt the synonymies, and, indeed, that establishment of such synonymies would be a disservice to araneology.
English-born John Abbot immigrated in 1776, as a young man, to Screven County, Georgia, and lived in Georgia for 65 years as a schoolmaster and naturalist. He painted birds, butterflies and other animals, and his drawings were sold by John Francillan, a London silversmith (Dow, 1914). Sixteen volumes of Abbot’s drawings are in the British Museum (Natural History) in London; many volumes
'We wish to thank the following for reading a draft of this manuscript and for making suggestions (without implying that they necessarily agree with the conclusions): Mrs. D. L. Frizzell (Dr. Harriet Exline), Dr. R. Crabill, Dr. C. Dondale, Prof. E. Mayr, and Dr. W. J. Gertsch. Dr. Gertsch kindly sent us a manuscript, prepared ten or twelve years ago but never published, in which he discussed the problem of the Walckenaer names. Although he believed the names were correctly synonymized by Chamberlin and Ivie (1944), he pleaded their rejection in the interests of nomenclatural stability. Dr. Gertsch and Dr. Dondale called our attention to pertinent literature. A National Science Foundation grant made possible our trip to Europe in 1958, at which opportunity we examined the Abbot manuscript drawings.
53
54
Psyche
[June- September
are in other institutions, several at the Houghton Library of Harvard University. According to Chamberlin and Ivie (1944), Walckenaer purchased drawings from the entomologist Mackay. In the eighteen forties Walckenaer named and described some of the drawings of the 1792 Abbot volume (now in the British Museum), in Histoire Naturelles des Insects Apteres. There is some doubt about the date of publication of Walckenaer’s second volume. Our personal volume has two inscriptions, one of Walckenaer, addressed to Mr. Adam White and dated 4 June 1841, and another presumably in White’s handwriting: “Adam White Villeneuve, St. Germ. Walckenaer’s study June 7, 1841.” Thus the publication date is undoubtedly 1841, not 1842 as stated by Chamberlin and Ivie and Bonnet (1945).
Dr. McCook visited the British Museum in 1887 and his atten- tion was called to the Abbot drawings of American spiders. In a report to the Philadelphia Academy of Sciences (1888a), he discussed some of the questions raised by this discovery. McCook was much concerned about the changing of names in use, but he also wanted to credit the earliest author: “the laws of priority must be con- sidered, and honesty and justice can give no room for considerations of convenience and sentiment.” Several argiopid names of Hentz were identified with those of Walckenaer.
McCook’s paper was reviewed by Emerton (1888). Emerton had looked over the Abbot drawings at the time of his visit to the British Museum in 1875, “and like Mr. McCook made hasty identifications of such few of them as I could. ... A comparison of the numbers shows that only five of these identifications agree with those of Mc- Cook showing the uncertainty of off-hand identifications of these drawings by two persons both familiar with the common spiders of the northern states. The greater number of Abbot’s drawings repre- sent the spiders only in the most general and indefinite way and it seems to me improbable that any large number of them can ever be identified.”
Included in McCook’s self defense (1888b) were excerpts from a congratulatory letter from Thorell. Banks followed: “The de- scriptions of new species in Walckenaer’s Insectes Apteres fall into two classes: descriptions based on specimens, and descriptions based on figures. The former class are undoubtedly valid and I intend to accept them wherever I can apply them. Descriptions of figures, however, I hold, have no claim on the naturalist. Not only are they based on figures, but the figures have never been published. Many of the descriptions are sufficient for identification, but most are not.
1961]
Levi and Levi — American Spiders
55
But no matter how complete, they are not descriptions of spiders; but of figures of spiders. They rank with ‘hearsay evidence’. I shall not use them nor list them ; I shall ignore them.”
Chamberlin and Ivie (1944) undertook “initially to determine, as far as possible from available evidence, the proper application of the names based by Walckenaer upon Abbot’s drawings of the spiders of Georgia.” Chamberlin made color photographs of Abbot’s drawings at the time of a London visit, and Ivie spent a month in April 1943 collecting spiders in Georgia. Some other collections were obtained during brief stops in Georgia in August 1933 and June 1935. Cham- berlin and Ivie listed the collections (including many determined juveniles), and synonymized many well established spider names of many families with names of Walckenaer.
However, in our own examination of the Abbot manuscript draw- ings, we found that the majority do not show diagnostic characters; interpretation must be subjective, and authors may differ. For in- stance, McCook synonymized the name T etragnatha lacerta Walck- enaer with T etragnatha caudata Emerton ; Chamberlin and Ivie synonymized the same name with Rhomphaea fictilium (Hentz), of a different family. Most of McCook’s synonymies concerned argiopid spiders that have a characteristic dorsal abdominal pattern. However a modern author has to consider the possibility of sympatric sibling species.
Further, and to be expected, students working with groups never revised make errors in identification. Thus Chamberlin and Ivie synonymized Argyrodes trigonum (Hentz) with Linyphia rufa Walckenaer. However, the specimens so labelled were not Hentz’s species, but were Argyrodes furcatus (O.P.-Cambridge) , a species more common in Georgia. Tidarren fordum (Keyserling) was syn- onymized with Theridion sisyphoides Walckenaer, but specimens so labelled were not Tidarren fordum. Female specimens of Theridion alahamense Gertsch and Archer were misidentified as Theridion amer- icanum Walckenaer, and the male was described as new. Walckenaer’s description of T. americanum does not fit T. alabamense. Figure 43, Walckenaer’s Theridion ansatum , was not recognized as the species otherwise called Tidarren sisyphoides (Walckenaer), easily recog- nized by the white line on the posterior part of the abdomen. Probably half the examined theridiids of the Chamberlin and Ivie Georgia col- lection had incorrect identifications, and the same may be true of specimens of other families. While these errors can easily be under- stood, they invalidate many of the synonymies of Walckenaer’s names.
56
Psyche
[June- September
Chamberlin and Ivie apparently were unaware of the earlier ex- changes about the synonymies, for they neither referred to them nor listed them in their bibliography. Also they apparently did not con- sider the possibility of additional plates, letters and notes concerning the Abbot spiders, in other libraries. Such sources of evidence might have relevance in interpreting Abbot’s drawings, which must be con- sidered the types for the Walckenaer names.
As recently as twenty years ago, some considered the discovery of an “older name” a matter to* be admired, and priority was frequently invoked to rationalize changing a name in widespread use. Today the attitude toward stability of names has changed, partly due to criticism of zoologists in fields other than taxonomy. Perplexed by the constant change of names, these other biologists leveled the charge that some taxonomists were spending more time in historical than in biological research.
Two approaches to stability of names present themselves: Some taxonomists think that through priority, the supply of older names will eventually become exhausted; at the same time, strict priority leads nomenclature back to the oldest and most uncertain names. Other taxonomists favor established usage of the name as the basis of stability, though usage may be hard to define.
The new (1961) Zoological code of Nomenclature combines the criteria of priority and usage through a statute of limitations. Also, and more important, the new code emphasizes in its Preamble the reason for its existence — to keep names stable — a reflection of the present needs of zoologists. The establishment of sweeping synonymies of the Walckenaer names based on Abbot’s drawings, so inconsistent with stability, should be questioned.
A more acceptable treatment of the Walckenaer names was demon- strated by Bishop (1924), in his revision of the North American Pisauridae. By restricting himself to one family he was able to treat each nomenclatural problem individually and authoritatively, rather than all in one arbitrary sweep. The largest number of specimens, the greatest grasp of the literature, and the keenest understanding of the particular spider group were brought into each judgement.
Our purpose, then, is to urge that spider students adopt the Walck- enaer synonymies proposed by Chamberlin and Ivie only after thor- ough study of the spider genera in question, including, in addition to a study of the Abbot drawings, investigation of usage of names, species problems, and distributions, giving due consideration to the basic principles of nomenclature : to stability and universality of names.
1961]
Levi and Levi - — ■ American Spiders
57
References
Archer, A. F.
1946. The Theridiidae or comb-footed spiders of Alabama. Paper Ala- bama Mus. Nat. Hist., no. 22: 1-67.
1950. A study of theridiid and mimetid spiders. Ibid., no. 30: 1-40. Banks, N.
1901. Notes on some species of Walckenaer, Koch and others. Jour. New York Ent. Soc., 9: 182-189.
Bishop, S. C.
1924. A revision of the Pisauridae of the United States. Bull. New York State Mus., no. 252: 1-140.
Bonnet, P.
1945. Bibliographia Araneorum, Toulouse, 1 : 832.
Chamberlin, R. V. and W. Ivie.
1944. Spiders of the Georgia region of North America. Bull. Univ. Utah, biol. ser., 8 (5): 1-267.
Dow, R. P.
1914. John Abbot of Georgia. Jour. New York Ent. Soc., 22: 65-72.- Emerton, J. E.
1888. Walckenaer’s names of American spiders. Psyche, 5: 113-114. Gertsch, W. J.
1933. Notes on American Lycosidae. Amer. Mus. Novitates, no. 693:1-2.
1953. The spider genera Xysticus, Coriarachne and Oxyptila in North America. Bull. Amer. Mus. Nat. Hist., 102: 415-482.
Levi, H. W.
1954. Spiders of the genus Euryopis. Amer. Mus. Novitates, no. 1666: 1-48.
Levi, H. W. and H. M. Field.
1954. The spiders of Wisconsin. Amer. Midland Nat., 51: 440-467. McCook, H. C.
1888a. Necessity for revising the nomenclature of American spiders. Proc.
Acad. Nat. Sci. Philadelphia, 1888: 74-79.
1888b. The value of Abbot’s manuscript drawings of American spiders. Ibid.-. 428-431.
Walckenaer, C. A.
1841. Histoire Naturelle des Insectes Apteres, 2, Paris.
THE NEOTROPICAL SPECIES OF THE ANT GENUS STRUMIGENYS FR. SMITH: MISCELLANEOUS CONCLUDING STUDIES
By William L. Brown, Jr.
Department of Entomology, Cornell University
This paper is a continuation of my series on the New World fauna of the dacetine ant genus Strumigenys Fr. Smith. Earlier parts, con- taining keys to the abbreviations for measurements and proportions, may be found in Jour. New York Ent. Soc. 61: 53-59, ioi-iio (I953)- In addition to these, other parts have been published in the same journal, in Psyche, and in Studia Entomologica, Petropolis, Bra- zil. This section is a final one so far as currently available material in the genus indicates ; only one more section, which will be composed chiefly of an illustrated key to the New World members of the genus, is planned at present.
S. micretes and S. lacacoca
Following the descriptions of what I called the emeryi group (Brown, 1959) the species were discussed as follows:
“The four species emeryi, never marnii , micretes and lacacoca are very close, and seem, from the limited material available, to replace one another in a chain extending from Mexico to Panama, and per- haps beyond. So far as I can see now, the differences are complex enough and strong enough to indicate that each form is a distinct species; perhaps together [they constitute] one superspecies. How- ever, it is not beyond possibility that one or more of these forms inter- grades with a neighbor. More material is needed.”
Since that writing, material has turned up which, though small in amount, tends to bridge the gap between S. micretes Brown and S. la- cacoca Brown, indicating perhaps that they belong to a single variable species. Nevertheless, the new material poses certain problems itself, and the discussion next offered is intended to give details that should help in eventually straightening this complex out.
A sample consisting of parts of four nest series from Boquete, Chiri- qui Province, Panama (F. M. Gaige leg., see below) contains 25 workers with highly variable preapical mandibular dentition, the den- ticles varying in number from 1 to 4 in the two mandibles taken to- gether, and also varying markedly in size, acuteness and position, so as to bridge virtually completely the chief diagnostic character-gap between micretes (each mandible with a small but acute preapical tooth, and a little farther up a minute denticle) and lacacoca (man-
58
1961]
Brown — Strumigenys
59
dibles completely without teeth or denticles basad of the apical fork in the preapical region). This sample indicates that the preapical dentition, usually so constant in species of Strumigenys , may some- times be unreliable. The Boquete sample would clearly indicate syn- onymy between micretes and lacacoca were it not for one disturbing fact: the Boquete sample differs from the available samples of both species in its larger size and in having the promesonotum very distinct- ly and closely longitudinally striate throughout (arched striate along the anterior pronotal margin). This sculpture is somewhat shining, especially on the sides, where a large section becomes smooth or nearly so. The rest of the sides of the alitrunk are also smooth and shining for the most part. Postpetiole predominantly smooth and shining, crossed by a few longitudinal costulae. Propodeal teeth also longer, more slender and more nearly horizontal than in the micretes or la- cacoca type series. Anterior coxae smooth or nearly so, shining.
Some specimens of micretes and lacacoca have feebly indicated longi- tudinal rugulae or costulae on the pronotum (in addition to the medi- an carinula), but in these the predominant sculpture is the usual opaque reticulo-punctulation over at least the discal portion. The size, head wddth and sculptural traits of the Boquete sample could well be diagnostic of still another species in this close-knit complex, or they could merely mark a local population of a single variable species that would also include the types of micretes and lacacoca. For the present, it seems wise to avoid introducing new species names for mem- bers of this complex and also to hold off from synonymizing micretes and lacacoca until the distribution and variation of the complex are better known. For the convenience of future workers, I list here the material of the complex that I have studied, with such measurements, proportions and other observations as I have obtained from them (n = number of workers measured for each sample) :
Colombiana Farm, Santa Clara, Costa Rica (W. M. Mann leg.), TL 2.9-3. i, ML 0.70-0.74, ML 0.45-0.47, WL 0.72-0.76 mm; Cl 71-74, MI 62-65 (n = 12), type series of S. micretes. Progreso, Chiriqui Prov., Panama, (F. M. Gaige leg., no. 332), TL 3. 1-3.3, HL 0.75-0.76, ML 0.50-0.51 mm; Cl “about as in the type series” of micretes, MI 66-68 (n = 6), series placed with micretes in the original description of that species. Boquete, Chiriqui Prov., Panama (F. M. Gaige leg., nos. 208, 497, 504 and one series with no number) TL 3.7-4.0, HL 0.86-0.90, HW 0.66-0.68, ML 0.58-0.61, WL 0.92- O.99 mm; Cl 75-76, MI 67-68 (n = 25), Strumigenys near micretes, discussed above. Cerro Campana, west of Chorrera, Panama Prov., Panama, at about 950 m altitude in montane rain forest (cloud for-
6o
Psyche
[June-September
est), in rotten wood (G. B. Fairchild and W. L. Brown leg., no. B-86), TL 3.3, HL 0.76, HW 0.52, ML 0.53, WL 0.82 mm; Cl 68, MI 70 (n — 1), specimen here placed as S. lacacoca; this worker has a single extremely minute denticle in the preapical concavity of the left mandible, perhaps representing a vestige of a distal preapical tooth. Rio Chinillo, Panama Canal Zone (T. E. Snyder leg.), TL 2. 8-3. 2, HL 0.70-0.75, ML 0.45-0.48, WL 0.72-0.75 mm; Cl 67-71, MI 60-66 (n 7), type series of S. lacacoca.
Though the known variation of the nvcretcs-lacdcoca is consider- ably expanded by the new material reported above, this variation is mainly away from the direction of S. nevermanni Brown, from Costa Rica, a species that is generally smaller, with a relatively shorter head and mandibles and narrow infradental lamellae.
A new species of the T ococae group
Strumigenys fairchildi new species
Holotype worker: TL 4.1, HL 0.96, HW 0.72, ML 0.59, WL 1. 01, scape L 0.68 mm; Cl 75, MI 61.
Aside from its larger size and relatively longer mandibles, this species closely resembles S. tococae Wheeler from the Amazon Basin. The upper propodeal teeth are somewhat longer (about as long as the distance between the centers of their bases, and about twice as long as the lower teeth), and the lamella between the upper and lower teeth is lower and more cariniform. The eyes are about the same size (greatest diameter about 0.14 mm), and the petiole and postpetiole are similar, but with slightly less voluminous spongiform appendages. The best distinction lies in the sculpture and pilosity of the nodes and gastric dorsum:
( 1 ) Postpetiolar disc convex, evenly reticulo-punctulate, with only feeble anterior traces of longitudinal costulae. First gastric tergite finely longitudinally striolate, opaque-sericeous throughout (in S. tococae , the gaster is costulate at the base, but otherwise1 smooth and shining) .
(2) Ground pilosity strongly reduced, virtually absent on nodes and gastric dorsum (plentiful and conspicuous in S. tococae). Fine erect flagelliform hairs of nodes and gaster shorter and apparently fewer than in S. tococae.
The color is ferruginous yellow, the gaster somewhat brighter yel- low than the rest of the body.
Holotype [MCZ] a unique worker from the wooded ravine next to the inn near the top of Cerro Campana, Panama Prov., Panama, 17 January, i960, at an altitude of 800 to 850 m (G. B. Fairchild
1961]
Brown — Strumigenys
61
and W. L. Brown leg.). The worker was found walking on a rotten stick lying on the leaf litter on the shaded slope of the ravine ; hasty search of the litter and soil nearby failed to uncover more specimens. Since the very closely related S. tococae was found at Belem, Brazil, nesting in foliar sacs of the plant Tococa, it is possible that S. fairchildi was also nesting in a plant cavity above the ground. Since the rela- tionships of S. fairchildi were not recognized until it was critically examined in the laboratory, the possibility of a plant-cavity habitat was not investigated in the field.
Group of S. lanuginosa Wheeler
Two species, S. lanuginosa Wheeler and S. hindenburgi Forel, share a number of traits that apparently indicate a fairly close relationship between them : Mandibles lying close together at full closure, their bases sharply narrowed from the outside; apical fork moderate in length, with a single intercalary tooth; inner (masticatory) margin concave near apex, the concavity with a short but acute preapical tooth ( S . hindenburgi has an additional minute denticle near the apical third of the margin). Clypeus broadly triangular, with convex an- terior margin. Antennal scape slender, tapered toward both ends, very slightly curved at basal third.
Pronotum with humeral angles developed, bluntly tuberculate; an- terior margin present (weak in S. lanuginosa) ; alitrunk in side view with convex dorsal profile, broken only at the impressed metanotal groove. Propodeal teeth of modest size, but acute, each subtended by a low, concave infradental lamella ending in a ventral convexity. Petiole distinctly pedunculate and with a short, dorsally rounded node; postpetiolar disc convex; both nodes with complete and well- developed spongiform appendages. Gaster normal in form, with a strong anterodorsal spongiform margin and a thick anteroventral spongiform pad ; basigastric costulae well developed.
Head, alitrunk and both nodes densely and finely reticulo-punctu- late, opaque; postpetiolar disc usually with weak rugulosity or costu- lation superimposed. Underside of gaster smooth and shining; mandibles weakly shining in some lights, punctulate; legs and an- tennae finely and densely punctulate. Sides of pronotum, anterolateral surfaces of anterior coxae, and lower lateral surfaces of infradental lamellae of propodeum with patches of fine reclinate hairs and fre- quently encrusted with whitish material; apparently these represent secretory areas.
Ground pilosity of head, promesonotum, posterior propodeun; scapes and legs consisting of fine, abundant reclinate and arched-rec.lin-
62
Psyche
[June-September
ate hairs. Larger specialized hairs all fine, long, flagelliform, more or less erect: a pair on each lateral occipital border (2 pairs in lanugin- osa)I, a pair on the middle occiput, a pair on the humeri and another pair on the mesonotum. Both nodes and both upper and lower sur- faces of the gaster with a luxuriant growth of abundant, long, fine flagelliform hairs, the longest of which, on the gastric dorsum, are as long as or longer than the petiole. Each inner mandibular border bears a row of fine oblique sensory hairs.
Males of neither species have yet been found. S. lanuginosa occurs in Central America and southern Mexico, and has apparently been introduced into the Bahamas, while S. hindenburgi is known from southern Brazil and northern Argentina.
Struinigenys lanuginosa Wheeler
Strumigenys lanuginosa Wheeler, 1905, Bull. Amer. Mus. Nat. Hist. 21: 104,
fig. M, worker, female. Type loc. : Fort Charlotte, Nassau, Bahama Islands.
Syntypes in AMNH, MCZ, several examined.
Worker: TL 24-2.7, HL 0.55-0.60, ML 0.35-0.38, WL 0.58- 0.64 mm; Cl 79-80, MI 63-64. Measurements from 5 workers, in- cluding a syntype, Panamanian and Mexican specimens.
Main distinguishing features:
( 1 ) Small size and slender build.
(2) Dorsolateral borders of head merely cariniform, not lamellate.
(3) Mandibles without a trace of a denticle basad of preapical tooth.
(4) Basal costulae of gaster extended as fine, sericeous striation (striolation) over the basal third or more of the first segment. Gastric pilosity also more abundant and crowded than in S. hindenburgi.
( 5 ) Two pairs of flagellate hairs on the lateral occipital margins.
(6) Postpetiole usually with traces of fine longitudinal striolation or costulation overlying the punctulation.
Color light ferruginous, gaster often slightly more brownish. Man- dibles and appendages lighter, more yellowish.
Female (dealate) : TL 2.9-3.0, HL 0.62-0.63, ML 0.36-0.38, WL 0.72-0.73 mm; Cl 80-84, MI 57-61 (from 2 syntypes). Males un- known.
Distribution: Southern Mexico, Panama; in Bahamas, where prob- ably introduced historically.
Localities for material examined: Bahama Islands, Nassau (W. M. Wheeler leg.), type locality. Mexico, Veracruz: Cordoba (C. H. Seevers leg.) , one worker under stone. Pueblo Nuevo, near Tetzonapa (E. O. Wilson leg.), strays from leaf litter in tropical evergreen for-
1961]
Brown — Strumigenys
6 3
est, and foraging on surface of log in degraded rain forest. Panama Canal Zone: Barro Colorado Island (J. Zetek leg.), a single worker. This species is here reported from the American mainland for the first time ; apparently it is widespread.
Strumigenys hindenburgi Forel
Strumigenys Hindenburgi Forel, 1915, Bull. Soc. Vaud. Sci. Nat., 50: 357, worker. Type loc. : “Argentine,” La Plata from label on syntypes. Syntypes in Coll. Forel, Museum d’Histoire Naturelle, Geneva, and MCZ, several examined.
Strumigenys hindenburgi , Kempf, 1958, Stud. Ent. (n.s.) 1:555, record from 10 km sw of Agudos, S. Paulo State, Brazil, among dry leaves in wooded gully.
Worker: TL 2. 8-3. 2, HL 0.60-0.69, ML 0.38-0.40, WL 0.63-0.71 mm; Cl 83-88, MI 58-64. Measurements from 3 syntypes and 6 additional Argentinian specimens representing 3 nest series.
Within the lanuginosa group, this species is distinguished by its large size and relatively broad head, as well as the following charac- ters :
(1) Dorsolateral borders of the head (dorsal scrobe borders) on each side produced as a narrow but distinct lamellar margin that grad- ually narrows posteriad and ends in a shallow concavity behind the level of the eye.
(2) Mandibles each with an extra minute preapical denticle, best seen in oblique view in dark silhouette against a bright background, near the apical third of the shaft.
(3) Basigastric costulae extending less than half the length of the first segment, the rest of which is smooth and shining.
(4) Lateral occipital margins each with only one flagellate hair, arising from the concavity at the end of the lamellate dorsolateral margin.
(5) Occiput, pronotum and postpetiolar disc with traces of feeble reticulate rugulation superimposed on the basic densely punctulate sculpture.
Color medium ferruginous ; legs, mandibles and antennae more yellowish ; gaster mostly brown.
Pseudogyne or ergatoid female from Itatiaia: TL 2.7, HL 0.60, HW 0.49, ML 0.34, WL 0.60 mm; Cl 82, MI 57. This small in- dividual has the mesonotum developed in the direction of the full female, with raised margins dorsad, and the differentiated scutellum acutely projecting posteriad. The petiolar and postpetiolar nodes are wide, as expected in females of this genus, but the gaster is not un- usually bulky for a worker. A tiny apparent remnant of a median ocellus occurs in the central vertex. T his specimen appears to me to
64 Psyche [June-September
be classifiable as a pseudogyne, or pathological worker-female inter- mediate.
Distribution: Northern Argentina, extending into southeastern Bra- zil.
Localities for material examined: Argentina: La Plata (C. Bruch leg.), 3 syntypes. Salinas, near Tucuman (Kusnezov and Golbach leg., no. 1677). Parque Avellanida, Tucuman (P. Wygodzinsky leg.), a small series of workers. Brazil: Rio de Janeiro State, Itatiaia, Lago Azul (R. Barth leg.), the single pseudogyne described above.
The Variation and Synonymy of Strumigenys louisianae Strumigenys louisianae Roger
Strumigenys louisianae Roger, 1863, Berlin, ent. Zeitschr. 7: 211, worker. Type loc. : “Louisiana.” Type in Zoologisches Museum der Humboldt Universita’t, Berlin, not examined
Strumigenys unidentata Mayr, 1887, Verh. zool.-bot. Ges. Wien 37: 575 and in key, p. 570, worker. Type loc.: “St. Catharina.” Lectotype, by present designation, in Naturhistorisches Museum, Vienna, examined, new syn- onymy.
Strumigenys fusca Emery, 1894, Bull. Soc. ent. ital. 26: 215, pi. 1, fig. 8, work- er. Type loc.: Manicore, Amazonas. Holotype in Museo Civico di Storia Naturale, Genoa, examined, new synonymy.
Strumigenys clasmospongia Brown, 1953, Psyche 60: 2, worker. Type loc.: Petropolis, Rio de Janeiro State, Brazil. Holotype in Coll. W. W. Kempf (ex Coll. T. Borgmeier), Sao Paulo, Brazil; paratypes in Coll. Kempf, USNM, MCZ, several reexamined, new synonymy.
In my “Revisionary Studies” of 1953, I showed that S. louisianae is a very variable species ranging from warm temperate North Ameri- ca south into Bolivia and northern Argentina. However, variation at that time was thought to involve mainly body size and proportions of the head and mandibles. After prolonged study, a number of names (see synoptic synonymy below) was placed in the synonymy of S. louisianae ; types were compared in most of these cases. The variety longicornis was also synonymized on the basis of its original descrip- tion.
In another paper entitled, “Three new ants related to Strumigenys louisianae Roger” (Brown, 1953a), I described S. clasmospongia , S. producta and S. mixta , stating that “each of the three forms has been compared with all other neotropical species known to me, both des- cribed and undescribed, except S. fusca and S. unidentata, two species to be placed among the species inquirendae .” Since that writing, I have been able to study the (previously unavailable) types of S. fusca, S. unidentata and S. unispinulosa var. longicornis on deposit in Vienna and Genoa, through the kindness of Dr. Max Beier and Dott. Delfa Guiglia respectively. From the first comparison, it was clear that my
1961]
Brown — Strumigenys
65
clasmospongia was only a larger, relatively narrow-headed variant of unidentata. The unique type of S. fusca was more puzzling, being a rather large louisianae- like worker with fairly broad head and longish mandibles, but having the posterior half of the first gastric tergite nearly completely smooth and definitely shining. Var. brevicornis turned out as expected: a large louisianae female fitting fairly well with the largest worker forms from South American samples of the species, so that my earlier synonymy was confirmed.
In addition to these types, I have been able during the last few years to examine an increasing accumulation of louisianae- related forms from southeastern Brazil and northern Argentina, and scattered specimens have even come in from central Brazil and Surinam, previously blanks on the map. These samples are due mostly to the kindness of Father Borgmeier and Father Kempf. Study of this new material has com- pletely changed my ideas on relationships of the species within the louisianae complex. First, the relatively slender, long-mandibulate forms with shining gastric dorsum were shown to be common in southeastern Brazil; it is these forms to which the names unidentata and clasmospongia have been applied. Taken in combination, the dis- tinctive <( unidentata characters” give the impression of a distinct species inhabiting southeastern Brazil, but all attempts to define a species with these characters have failed. The reason for this failure is that the new material clearly shows that each of the characters go- ing to make up the unidentata in its “typical” manifestation actually has its own independent pattern of geographical and individual varia- tion within louisianae. The long mandibles (MI up to 67) of the southeastern Brazilian samples are approached by samples from Bo- livia (MI 60-63), and some series from southeastern Brazil have much shorter mandibles (MI as low as 56, and perhaps even lower in some samples seen but not measured), in this respect being not far from average for the species louisianae taken as a whole. The gastric sculpture of the unidentata pattern, in which basigastric costulae of moderate length are followed by a smooth or nearly smooth, shining surface (though this surface may be obscured by secretion or other foreign matter), is not confined to southeastern Brazilian specimens with long mandibles; in fact, it is found in series with much shorter mandibles from localities as far away as Tucuman, in northern Ar- gentina; Goias, Amazonas and Surinam in the northwest and north; and even in southern Mexico (Veracruz, Puebla). Furthermore, cer- tain series even in southeastern Brazil contain workers with shorter mandibles, some individuals of which have the reticulate, opaque " louisianae sculpture”, while others from the same nest sample have
66
Psyche
[June-September
the gaster predominantly smooth and shining. Variation in the queens is poorly known because not many of the “ unidentata” workers are accompanied by females in the collections I have seen. In general, loaisianae- complex queens have stronger gastric sculpture than the workers accompanying them.
In view of the discordant nature of the variation in the only good distinctive characters available, I am forced to consider fusca, uniden- tata, and clasmospongia as synonyms of louisianae. It is possible that the variation of this very plastic species is even greater in central and northern South America, from which our samples are so few, and per- haps even the large, very long-mandibulate producta is only another extreme variant of louisianae. The type of fusca does show tendencies in the direction of producta, but we shall need more material from Western Brazil and Bolivia before we decide this question. Of course, the possibility must not be overlooked that louisianae really is made up of a number of cryptic species, inseparable by conventional mor- phological study.
A fact of continuing interest is the absence of S. louisianae from the forest on Barro Colorado Island in the Panama Canal Zone. Inten- sive collecting by a number of mvrmecologists on the Island was re- peated in January i960 by Dr. E. S. McCluskey and myself, making full use of Berlese funnels and other modern collecting techniques, but no one has yet found 5. louisianae on the Island or elsewhere in Panama. This is especially strange in view of the fact that the species is common in banana plantations on both the Atlantic and Pacific sides of Costa Rica near the Panama border (El Palmar and Coto in the Golfo Dolce, E. O. Wilson leg.). While we now have very inade- quate ecological information, it does seem possible that S. louisianae may be a species that has adapted to habitats marginal to the rain forest of the South American continent, and that this has something to do with its present wide distribution — the widest of any New- World dacetine. In this sense, S. louisianae may fit Wilson’s (1959) “Stage-I” category of expanding species. It is also of interest to note that the species is much less variable (“more typical”) in the North American extremities of its range than in the presumed evolutionary center in South America. Furthermore, the “typical” characteristics of short mandibles and reticulate gastric sculpture, while discordant one with the other geographically, tend to prevail at the extremities of the range in North America and South America as well, indicating a centrifugal evolution and movement of these characters.
Belowr I have listed some of the available samples of S. louisianae by geographical regions, with special emphasis on some of the more
1961]
Brown — Strumigenys
67
significant South American samples reviewed since 1953, and giving certain measurements and proportions as known for the combined samples from each region (n = number of workers in each combined sample) .
Southeastern United States (8 localities in 5 states) : HL 0.52-0.61, ML 0.27-0.23 mm; Cl 82-87, MI 52-57 (n = 15). Guatemala and Costa Rica: HL 0.50-0.58, ML 0.27-0.33 mm; Cl 83-85 (n = 11). Cuba and Puerto Rico: HL 0.52-0.60, ML 0.26-0.32 mm; Cl 83-87, MI 51-56 (n = 11). Colombia (Rio Porce and Medellin, leg. N. A. Weber) : HL 0.54-0.61, ML 0.32-0.34 mm; Cl 71-84, MI 55-56 (11 = 3). Surinam (La Poulle and Dirkshoup, leg. I. van der Drift) : HL 0.49-0.53, ML 0.29-0.30 mm; Cl 83-86, MI 57-59 (n = 2). Goias (Anapolis, leg. W. W. Kempf) : HL 0.47, ML 0.23 mm; Cl 85, MI 49 (n = 1). Southeastern Brazil (States of Rio de Janeiro, Sao Paulo, Santa Catarina and Rio Grande do Sul) : HL 0.46-0.58, ML 0.28-0.37 mm ; Cl 77-92, MI 56-67 (n ^ 22). Bolivia (Rosario, leg. W. M. Mann) : HL 0.50-0.53, ML 0.31-0.32 mm; Cl 81-83, MI 60-63 (n — 6). Holotype female of var. longicornis from Coroi- co, Bolivia: TL 3.0, HL 0.66, ML 0.38, WL 0.77 mm; Cl 85, MI 58. Northern Argentina (3 localities) : HL 0.49-0.64, ML o. 26-0.35 mm; Cl 83-89, MI 52-56 (n = 6).
An additional record of interest is a sample of 5. louisianae from Yuma, Arizona, “on cotton,” “HAGA” leg.
Descriptive Note on the Holotype of S. fusca
Holotype worker: TL 3.4, HL 0.67, ML 0.40, WL 0.71 mm; Cl 82, MI 59. A large variant of S. louisianae ; mandibles farther apart and relatively more slender than in North American louisianae samples ; inner margins not so sharply concave near apex. Apical fork : dorsal tooth about 0.12 mm long, ventral about 0.09 mm; two sub- equal intercalary denticles. Preapical tooth about 0.05 mm long, and its tip situated about 0.05 mm from dorsal apical tooth ; slightly curved toward mandibular apex; larger and farther from apical fork than in most S. louisianae. No other teeth or denticles on inner mandibular border. Scape L 0.44, funiculus L 0.58 mm.
Promesonotum high, with convex profile, promesonotal suture (sul- cus) visible in some lights. Metanotal groove weak. Propodeal teeth acute, elevated, with narrow infradental lamellae approaching carini- form; teeth about 4/5 as long as the distance between their basal cen- ters. Petiolar peduncle longer than node; node broader than long. Postpetiole robust, fully punctulate, opaque. Spongiform appendages as usual for S. louisianae. Gastric costulae extending about 1/2 the
68
Psyche
[June-September
length of the basal tergite, interspersed with reticulo-punctulation ; remaining half of tergite nearly completely smooth, distinctly shining. Pilosity as usual in S. louisianae, but a little coarser than the average. Color dark ferruginous, but not as dark as some specimens of the “ unidentatci” conformation from the wet Serra do Mar, the coastal mountain strip of Sao Paulo ; these latter samples approach black.
The fusca type is similar to producta as well as to more “typical” louisianae , but differs in details of cephalic proportions and in sculp- tural details. The body of the fusca type is more robust, and the hairs of the pilosity are larger throughout. The preapical tooth is larger and is situated a little farther from the apical fork.
Descriptive Note on the Lectotype of S. unidentata
Lectotype worker: TL 2.2, HL 0.55, ML 0.36, WL 0.54 mm; Cl 79, MI 66. Despite its slightly smaller size and wider head, there is little doubt that this type and the species I described as clasmospon- gia are conspecific. I have partially cleaned and reexamined the postpetiole and gaster in the available types of unidentata and clas- mospongia, and the postpetiole is now seen to vary from weakly to moderately shining discad, while the gastric dorsum now appears as smooth and shining, with reticulation nearly or quite absent, except for the basal costulae. The frequent presence of a refractory hardened secretion (?) on the gastric dorsum is misleading when describing sculpture in a number of dacetine groups. The specimen now in the Naturhistorisches Museum in Vienna has been selected as lectotype, and is so labeled.
Synoptic Synonymy of Stru/nigenys louisianae
In the list below are found the names currently considered to be synonyms of S. louisianae. Full page references for each synonym are to be found in Brown, 1935b, p. 28, or, in the case of new synonymy, with the species heading above.
S. louisianae Roger, 1863
= unidentata Mayr, 1887, n. syn.
^ unispinulosa Emery, 1890.
— unispinulosa var. longicornis Emery, 1894.
— fusca Emery, 1894, n. syn.
— louisianae var. obscuriventris Wheeler, 1908.
= bruchi Forel, 1912.
— inpdelis Santschi, 1919.
— eggersi var. cubaensis Mann, 1920.
= louisianae subsp. laticephala M. R. Smith, 1931.
— louisianae subsp. soledadensis Weber, 1934.
= louisianae subsp. guatemalensis Weber, 1934.
— louisianae subsp. costaricensis Weber, 1934.
= clasmospongia Brown, 1953, n. syn.
1961]
Brown — Strumigenys
69
References Cited
Brown, W. L., Jr.
1953a. Three new ants related to Strumigenys louisianae Roger. Psyche, 60: 1-5.
1953b. Revisionary studies in the ant tribe Dacetini. Amer. Midi. Nat., 50: 1-137, cf. pp. 28-31.
1959. The neotropical species of the ant genus Strumigenys Fr. Smith: group of emeryi Mann ( Hymenoptera ) . Ent. News, 70: 97-104. Wilson, E. O.
1959. Adaptive shift and dispersal in a tropical ant fauna. Evolution, 13: 122-144.
A NOTE ON THE ANT GNAMPTOGENYS HARTMANI WHEELER. — Workers and winged females of this ant were received recently from Dr. M. R. Smith of the U. S. National Museum and were determined by direct comparison with the type in the Museum of Comparative Zoology. The sample represents a nest taken in the soil of a banana plantation on Zapote Farm, La Lima, Honduras, May 18, 1961 (E. Molendez leg.). This is the first record of the species to come to light since the original description, published in 1915, based on a single worker specimen fromi Huntsville, Texas (C. G. Hartman leg.). The type locality in northeastern Texas has always seemed anomalous for a genus otherwise confined to a more strictly tropical climate and not known from any other samples occurring natively within the continental United States, but until the present find, the good possibility remained that G. hartmani was an extralimital relict. It now seems more likely that the Texas Record represents either a locality error or an adventive specimen taken from bananas. No ecological data accompanied the original find. At least, the occurrence of this (or other) species of Gnamptogenys in Texas remains to be convincingly demonstrated. — W. L. Brown, Jr., Department of Entomology, Cornell University.
ANTHICUS TOBIAS MARSEUL,
ANOTHER TRAMP SPECIES (COLEOPTERA: ANTHICIDAE)
By F. G. Werner University of Arizona, Tucson
Among the rather numerous collections of Anthicidae examined by the author during the past ten years there have been small numbers of an unidentified species of Anthicus from a truly amazing variety of localities. In the United States, which yielded the first examples seen, specimens have turned up from the Atlantic to the Pacific and from New England to Florida, including such places as Catarina, Texas and the Sierra Ancha Mountains of Arizona. The author must admit that he held these specimens to represent an undescribed species and had prepared a description for publication and designated types. Then a specimen was sent from Guam, in the Marianas, others from Oahu identified as Anthicus mundulus Sharp, and, most recently, others from Jamaica and Venezuela. By this time it had become apparent that the species must have been introduced into part of the localities. Structurally, it is quite unlike any other species in the New World, especially in the details of the very elaborate structure surrounding the male primary gonopore. So an Old World source was likely.
The true identity of the species was revealed in the British Museum collection, which the author was privileged to study recently. Speci- mens identified as Anthicus tohias Marseul from Aden, India and Java are the same as those previously seen. The type specimen of tohias has not been examined but the specimens at hand agree per- fectly with the original description. Both Anthicus mundulus Sharp and A. cervinus LaFerte, under whose names most specimens have been misidentified, are quite different.
In the United States, at least, this is not a very abundant species. In most cases only one or a very few specimens have been seen from any one locality. It may be more abundant on Oahu, if one can judge from the number of specimens that have accumulated in collections. In order to help prevent further misidentification in any new areas where the species may be found, a brief description and figure are given here. The male genitalia are very distinctive. In most cases an examination of the tip of the tegmen will suffice for the identification of the species; none other has been seen with a similar conformation. A quick perusal of the British Museum collection and the Pic collec- tion in Paris failed to disclose any species that might be likely close
70
1961]
W erner — Anthicus tobias
71
relatives, at least on external features. However, the author is not thoroughly familiar with the Old World fauna and may have missed some species.
Anthicus tobias Marseul (Figs. 1-3)
Anthicus Tobias Marseul, 1879, L’Abeille 17: 125. Pic, 1911, Coleop. Cat., pars 36: 77.
Rufous, the antennae, palpi and legs luteous; elytra with a ferru- gineous to piceous area that extends from the basal 2/5 to 1/4 to the apex, with the suture pale anteriorly, and with a very pale spot on each elytron at about 1/5 from the apex. Prothorax and elytra flattened.
Head subquadrate, but base rounded and slightly notched at the middle. Surface smooth, with dense, rather small punctures through- out, even on the midline. Pubescence moderately dense and short, fine, decumbent. Eyes moderately large, prominent. Antennae slender but slightly thickened apically. Prothorax widest at apical third, flattened, with punctures and pubescence similar to those of head. Elytra subparallel, flattened; omoplates slightly elevated; suture elevated on apical half. Surface smooth, moderately densely punctured, the punctures on the basal portion slightly larger and feebly asperate, those behind fine and small. Pubescence golden in the pale zone, brown over the dark markings, decumbent, fine, directed obliquely laterally in the basal fourth, not conspicuous. Tactile setae very short but erect. Male pygidium flattened, with truncate apex; hypopy-
.Mk
Figs. 1-3. Anthicus tobias Marseul. 1. Habitus sketch, with sculpture and pubescence omitted. 2. Aedeagus, in ventral view as it lies in the abdomen. 3. Aedeagus, in left lateral view.
72
Psyche
[June-September
gium with a shallow, V-shaped notch. Aedeagus with a dorsal ridge and subapical notch on the tegmen; gonopore armature very large and complex, not inverted.
The total length of the specimens examined ranges from 2.6 to 3.1 mm. Both the pale area at the base of the elytra and the subapical spots vary somewhat in size. The most similar-looking species in the North American fauna is Anthicus cervinus , which has curved pubes- cence on the elytra in most parts of its range, elytra that are not flattened, and very different male genitalia.
Range: Except for the localities followed by a bibliographic cita- tion, the author has seen specimens from each of the following locali- ties: ASIA: Arabia (Pic, 1911) ; W. Aden Prot. : Lahej — XI-27- 1937 - — - Scott & Britton; Iraq (Marseul, 1897) >* Turkestan (Pic, 1 9 1 1 ) ; India (Pic, 1911) and U. P. : Fyzabad — Hingston. IN- DIAN OCEAN: Mauritius (Pic, 1911). E. INDIES: Java: Pre- ange-Tijembong — 4-A5 — Corporaal. PACIFIC OCEAN : Mari- anas: Guam: Agana - — X-io-1952 — J. W. Beardsley; Hawaiian Is- lands: Oahu: Ewa — X-1958 — It. trap; Waipio — - I-22-1946 and
VIII-1949 — It. trap — Pemberton; Manoa — IX-10-1949 — at It.
— O. H. Swezey. NORTH AMERICA: CANADA: Quebec: Granby — VI-15-1911 — P. E. Mercier. U. S. A.: MASS.: Boston
— VIII-13-1911 — found dead on Boston Common — Fall Coll.; Bedford — VII-15-1911 — C. A. Frost; Nahant — VI-3-1935 — P. J. Darlington; Framingham — - VIII-12-1944 — C. A. Frost; Dover — VIII-3-1949 — K. Christiansen. CONN.: Hamden —
VIII- 25-1939 — It. trap — N. Turner. WIS.: Wood Co.: Nekoosa
— VIII-23-1948 — It. trap — W. W. Barrett. IFL. : Chicago —
IX- 6-1909 — Fall Coll, and X-24-1921 — A. B. Wolcott; LaSalle Co.: VIII-24-1936 — F. Werner, and Lowell — VIII-24-1948 — It. trap — F. Werner & W. Nutting. PENN. : Delaware Co. — IX- 10-1941 — R. C. Casselberry; Williamsport — VIII-3-1949 — at It. D. C. : Washington — VIII-15-191 1 — at It. — H. S. Barber. FLA. : Jacksonville — VI-10 — R. L. Blickle; Broward Co.: Pineland —
X- 1956 — L. N. Bell. ALA. : Mobile — VI-23-1950 — E. O. Wil- son. TEX.: Dimmit Co.: Catarina — VII-7-1948 — at It. — F. Werner & W. Nutting. ARIZ. : Sierra Ancha Mts. — X-1-1925 — D. K. Duncan. ORE.: Jackson Co.: Ashland — Black & Davis. W. INDIES: Jamaica: Spanish Town — flying at dusk; Morant Bay; and Gordon Town, all II- 1 937 — E. A. Chapin and R. E. Blackwelder. S. AMERICA : Venezuela; Caracas — V-VI-1957 — M. J. & S. Sargent.
CHEMICAL AND BIOLOGICAL CHARACTERIZATION OL VENOM OE THE ANT SOLENOPSIS XYLONI McCOOK
By Murray S. Blum1, J. E. Roberts, Jr.2, and A. F. Novak3
The reaction of human beings to the sting of the indigenous southern fire ant ( Solenopsis xyloni McCook) is in marked contrast to that produced by the sting of the closely related imported fire ant ( Solenopsis saevissima [Fr. Smith]). Whereas the sting of S. saevis- sima is characterized by a painful edema and marked necrosis (Caro et al. [1957]), we have found that the sting of S. xyloni seldom results in more than a mild prurience. These facts strongly indicate that the venoms of these two species of Solenopsis differ chemically. The purpose of this present paper is to compare the chemical and biological properties of these Solenopsis venoms in order to possibly determine what is responsible for their different dermatological effects.
Materials and Methods
Venom was collected from major or media workers employing a previously described method (Blum et al. 1958). The chemical and biological properties of S. xyloni venom were studied by procedures described elsewhere (Blum et al. 1958; Blum and Callahan i960). A crystalline derivative of the main component in S. xyloni venom was prepared from an ether extract of 450 poison glands dissected from major workers. The derivative was isolated by the method of Blum and Callahan (i960).
The dermatological effects of the sting of S. xyloni to human beings were studied by observing reactions at sting sites.
Results and Discussion
The chemical properties of the venom of S. xyloni parallel those of the venom of S. saevissima in nearly all respects. Like the venom of S. xyloni , the venomous secretion of S. saevissima consists of an alkaline two-phase system in which the suspended droplets represent the minor phase (Blum et al. 1958). The main constituent in the venom of S. xyloni is an amine which is chemically comparable to the amine isolated from the venom of S. saevissima (Adrouny et al. 1959; Blum and Callahan i960). The infrared spectrum of the venom of S. saevissima is virtually superimposable on the spectrum of the venom of S. xyloni and it is probable that the amine constituents which these
department of Entomology, Louisiana State University.
department of Entomology, Louisiana State University. Present Address: Louisville General Hospital, Louisville, Kentucky.
department of Agricultural Chemistry and Biochemistry, Louisiana State University.
73
1961]
Blum, Roberts , and Novak — Solenopsis
74
spectra represent are very similar. On the other hand, whereas the venom of S. saevissima contains two rhodamine-complexing minor components, the venom of S. xyloni contains only one.
The chemical similarities of the two venoms are paralled by their biological properties. The venomous principles of S. xyloni exhibit the same antimycotic and antibacterial activities as are found in the venom of S. saevissima (Blum et al. 1958). The pronounced hemolytic effect and insecticidal activity of S. xyloni venom compare to these same properties in the venom of S. saevissima (Adrouny et al. 1959; Blum et al. 1958). Thus the venoms of both of these fire ants feature the same broad-spectrum activity against diverse types of cells.
The skin responses of human beings to the stings of these two fire ants are similar only during the first few hours, both being charac- terized by an immediate flare followed by a wheal. However, whereas the sting of S. saevissima is always characterized by an umbilicated pustule at the sting site (Caro et al. 1957), we have found that the response to the sting of S. xyloni seldom results in more than a mild prurience. In the few cases where minute pustules were observed, they were on individuals who were quite sensitive to the sting of S. saevissima. At least three explanations seem possible : ( 1 ) minor structural modifications of the necrotoxin in the venom of S. saevissima are associated with a large increase in necrotoxicity when compared to its counterpart in the venom of S. xyloni (2) the concentration of the necrotoxin in the venom of S. saevissima is greater than its counter- part in the venom of S. xyloni, (3) the minor components contribute to the necrotoxic action of the venom. These hypotheses remain to be determined experimentally.
References Cited
Adrouny, G. A., V. J. Derbes, and R. C. Jung.
1959. Isolation of a hemolytic component of fire ant venom. Science 130:449.
Blum, M. S. and P. S. Callahan.
1960. Chemical and biological properties of the venom of the imported fire ant ( Solenopsis saevissima var. richteri Forel ) and the isola- tion of the insecticidal component. XI Int. Kongr. Ent., Vienna. 3:290-293.
Blum, M. S., J. R. Walker, P. S. Callahan, and A. F. Novak.
1958. Chemical, insecticidal and antibiotic properties of fire ant venom. Science 128:306-307.
Caro, M. R., V. J. Derbes, and R. Jung.
1957. Skin responses to the sting of the imported fire ant ( Solenopsis saevissima) . A. M. A. Arch. Dermotol. 75:475-488.
MASS INSECT CONTROL PROGRAMS:
FOUR CASE HISTORIES*
By William L. Brown, Jr.
Department of Entomology, Cornell University
PREFACE
Insect control is a vast subject. It encompasses many methods of approach meant to protect a wide diversity of human resources, in- cluding the lives and health of humans themselves. Upon the success or failure of insect control programs have rested the fate of armies, of great canals and populous lands. Yet, though man has registered many practical successes against particular insect menaces, we do not yet understand fully the underlying dynamics of insect populations (or for that matter, of other animals, including man himself), and until we do, perfect control will probably continue to elude us in many cases.
However, there exist practical measures that have been used suc- cessfully to control or eradicate many kinds of insects, even though
Figure 1. Insecticide sales by U. S. producers in recent years, projected through to the end of 1961. Domestic consumption of insecticides actually declined slightly during 1960 in the U. S., but exports more than made up this dip. From Chemical Week, July 22, 1961, by permission.
*This study and the report were sponsored and supported by the Conserva- tion Foundation, New York.
75
76
Psyche
[June-September
we may not understand exactly how a particular measure takes its effect. In recent years, developments in practical insect control have come thick and fast, particularly in the field of pesticides. The de- velopment since World War II of chlorinated hydrocarbons, carba- mate and organic phosphate insecticides, distributed by mass aerial spray techniques, has revolutionized control work and has raised insec- ticide production and aerial application to the status of big businesses. But, promising as it seemed in the immediate postwar years, simple mass aerial broadcasting of toxic materials has not always led to efficient control of the target pest. Furthermore, the extensive application of this relatively unselective technique inevitably caused damage to in- cidental targets — plants and animals or property valued by humans — and there even arose a threat to human health itself.9* 20 As such damage and threat of damage became more obvious, protest against mass air-spraying increased in volume, and naturally the demand grew for research into alternative means of control.
It is my intention now to attempt to illuminate the current status and outlook of insect control methods in the United States by out- lining four case histories of large-scale insect control programs. It is difficult to say how representative these case histories may be, considering the very diverse nature of insects and the damage each kind does. All four of the, programs are large and expensive ones as such operations go, all have been considered to be eradication programs at one time or another, and all have been guided or conducted by agencies of the United States Department of Agriculture (hereinafter referred to as USDA).
Since these great programs affect or involve many people and many diverse vested interests, they are all to some extent controversial. Because controversy about them involves many contradictory findings and interpretations, it is often difficult to gain a true and unbiased conception of what is going on in a given instance. For this reason, I have tried to draw my information from as large and varied a group of sources as I could find (see Acknowledgements and References Cited) . Let us now see if a resume of four programs — Gypsy Moth, Fire Ant, Mediterranean Fruit Fly and Screwworm — will help us to appreciate the problems of mass insect control.
THE GYPSY MOTH Introduction
The Gypsy Moth, Porthetria dispar (formerly hymantria dispar ), is a variable insect, a native of Eurasia, where it ranges from Portugal and North Africa to Japan. The insect was imported to the Boston
j
1961]
Insect Control Programs
77
area from France in 1869 by a misguided naturalist who believed that he could cross it with silkworms. Moths escaped from his breed- ing colony, but it was not until 1889 that the first severe outbreak defoliated fruit and shade trees in many towns of eastern Massachu- setts. Control work was started by the state and apparently was successful, for populations were so low by 1899 that control operations were ended. The moth soon again built up extensive populations, and control work was resumed in 1905, but it had spread by this time to western Massachusetts and parts of Maine, New Hampshire and Rhode Island. In 1906, Congress voted aid to the infested states to help prevent the spread of the moth, but despite all efforts it con- tinued to expand its range.
Biology and Nature of the Damage
The gypsy moth has a single generation per year. The winter is passed in the egg stage, and in New England the larvae hatch in mid- spring and feed through May and June, entering the quiescent pupal stage in early July. The larvae feed on a wide variety of broad-leaved trees and shrubs, especially oak, willow, poplar, birch, fruit trees and, in heavy infestations, even hemlock and pine. Dense populations may completely defoliate large jireas of forest, weakening many trees and killing others outright.
The heavy-bodied female does not fly, but puts out a powerful scent to which the strorTg-flying male responds, even to extremely minute amounts carried on the air great distances, by flying upwind until contacting the source individuals and copulating with them.18 The female deposits her eggs on tree trunks, fences, rocks and other solid objects. The young larvae spin silken threads on which they are easily spread by the wind before they start to feed.
According to Campbell4 the strong fluctuations in abundance of the moth are density-reactive, a most critical factor in this reactivity being the larval behavior. At low densities, the caterpillars tend to descend to the leaf litter to rest during the daytime, and feed mainly at night out on the foliage. When density is intermediate, the larvae rest during the day under loose bark on the tree trunks, a habit that has been used to advantage in control work (bands of burlap placed around trunks of infested trees are removed daily and the caterpillars found beneath them are destroyed). At high densities, the larvae remain on the foliage day and night, and are subject to heavy losses due to disease, desiccation and attack by ichneumon-wasp parasites. Population “crashes” are correlated with previous high densities of larvae.
78
Psyche
[J une-September
Control Problems
Early control efforts by the State of Massachusetts and the Federal Government included laborious and expensive methods such as hand- creosoting of egg masses, shelter-band and tanglefoot trapping on tree trunks, and various kinds of spray operations from the ground. For many years, control and quarantine programs appear to have confined the infestation to the area east of the “barrier” at the Berkshires and Green Mountains. Occasional extralimital infestations appearing in New Jersey, Ohio, Pennsylvania and Canada, particularly after egg masses were spread widely by the hurricane of 1938, apparently were eradicated before getting out of hand. Extensive introductions of predatory and parasitic insects from Europe and Japan were made beginning in 1905, and about ten such insects have taken hold in North America. Much of the subsequent history of the infestation was summarized in the report of the Gypsy Moth Eradication Meeting11 held in Ithaca, New York, in September, 1957:
“Following World War II, DDT was found to be a specific insecticide for the gypsy moth. At about the same time applica- tion of insecticide by plane became a practical undertaking. It was a new day for gypsy moth control. Heavy infestations within the area of general spread were suppressed or brought under control, and new infestations beyond the barrier were detected and held in check. Pennsylvania eradicated with reason- able effort and expenditure the gypsy moth on an area of 300,000 acres. Unfortunately more than 20 million acres were infested in this country before a practical control was discovered.
For some unexplained reason, the gypsy moth infestations seemed to explode* in 1950 and there was rapid spread beyond the bar- rier zone. Following the outbreaks in 1953 and 1954, surveys revealed the new areas of infestation west of the barrier zone in New York, New Jersey and Pennsylvania, aggregating nearly 9 million acres. An isolated infestation found in the vicinity of Fansing, Michigan, was immediately scheduled for eradication. The occurrence of these infestations west and south of the barrier posed a serious threat of spread to the hardwood forests through- out the eastern and southern United States. The control and quarantine programs that had successfully held the moth in check for so long were no longer adequate. ...”
*The explosion might better be said to have fairly begun in 1951 or 1952; see Figure 2. Its inception so soon after mass air spraying of DDT began on an operational basis is a phenomenon which, curiously enough, seems to have attracted little attention. It was first pointed out to me by Prof. F. M. Carpenter of Harvard University. — W. L. B.
1961]
Insect Control Programs
79
ACRES SPRAYED BY AIR
Figure 2. Graphs to show the ups and downs of the struggle against the gypsy moth in the U. S. Acreage showing substantial defoliation by gypsy moth larvae each year (below) is compared with acreage sprayed from the air (above) mostly with DDT at 1 lb per acre. Some suppression treatments used only 1/2 or 3/4 lb of DDT per acre, and sevin has partly replaced DDT in recent years. For details, see summaries by USDA in Appendix A, upon which these graphs are based.
8o
Psyche
[June-September
In spite of the difficulties involved, Federal and some state authori- ties were still speaking in terms of “eradication” of the gypsy moth in 1956 and 1957, while other state and local people were by this time hesitant about backing an all-out eradication effort.
In 1957, after about three and one-half million acres had been sprayed (two and one-half millions of them in New York State), DDT residues were found on forage crops and in the milk of cows that had grazed on treated areas in New York State, as well as in eggs from poultry farms that had received spray.16 DDT tolerances for milk are set at zero by the Federal Food and Drug Administration and by health authorities in New York among other states.
When the DDT residues were found persisting on forage crops and in the raw milk for periods up to a year, New York suspended eradication efforts “. . . so that,” as the USDA’s Cooperative Plant Pest Control Programs for 1958 put it, “the 1957 work could be fully evaluated and any required ‘mopping up’ could be done; how- ever, during the eradication season tests were made of several alternate insecticides more suitable than DDT for use on pasture and forage crops.”
Since 1958, New York has been doing a greatly reduced amount of spraying by air, using in part the new insecticide sevin, a carbamate having very low toxicity to mammals and birds, and one leaving no residue in the milk. Unfortunately, sevin is not as good against the gypsy moth as is DDT, it is highly toxic to honeybees, and it injures plants to some extent.
Aside from the dairy-linked residue problem, DDT has received rather good marks from most biologists checking the general ecological effects of mass spray at one pound to the acre. A few fish, are some- times killed, birds that catch insects on the wing depart, and certain aquatic insects suffer, but the known damage does seem tolerable. Long-term residual effects on soil organisms are, however, not well known.
The chief short-range danger of mass aerial DDT campaigns lies with the loose spray practices or accidents that result in duplication (or worse) of spray strips in a given area. Field insect control men often complain about the quality of pilots available for some spray programs, and numerous incidents have occurred to illustrate the point that some of the pilots are irresponsible or incompetent, or that they are poorly directed. For this and other reasons, it seems certain that operational mass spraying does not always give the same safe results as are found for the neatly-sprayed test strips of some of the studies, and landowners are often justified in complaining of double or triple
1961]
Insect Control Programs
81
doses of spray on their land. In view of these difficulties, DDT must be considered as only a marginally safe compound even at the I lb per acre dosage.
The issue of mass spraying has come to one court battle that at- tracted considerable attention. A group of plaintiffs led by Dr. Robert Cushman Murphy, the well-known ornithologist, sought injunctions against mass spraying of DDT for gypsy moth on or near their land, which was situated near New York City and mostly on Long Island. Most of the plaintiffs were organic gardeners and nature-lovers, and much of their testimony tended to be emotional in tone but rather insubstantial as to verifiable facts. The government defended itself with toxicologists and entomologists who presented a generally factual picture, and the case was decided against the plaintiffs by the Federal judge, although he warned the government to use more care in spray operations. The main effect of the case appears to have been to make the spray agencies hesitant about treating Long Island and many other farm areas. Also, by agreement with New York health authorities, a wide belt is left unsprayed around the large reservoirs of the metro- politan water supply. Such areas can of course provide refuges for the moth from which it is potentially able to recolonize adjacent treated areas.
Thus, for various reasons, the large key “border state” of New York has in fact been forced to abandon the “eradication” campaign, and the Plant Pest Control Division of the USDA now speaks instead of a “containment program” which would include chemical treatments within the infested area and along its periphery to back up the con- tinued quarantines.
Infestations in Pennsylvania and Michigan, thought on several past occasions to have been eradicated or nearly so bv DDT spray, still survive. Directly menaced are the hardwood forests of the Atlantic Slope, the Appalachians and the Mississippi Valley.
What Can Be Done About the Gypsy Moth?
I gather from conversations and correspondence with entomologists and foresters responsible for gypsy moth control at the state and local level that they generally share an uneasiness about the use of air- sprayed non-specific poisons such as DDT and sevin on forest and watershed areas. Most of them expressed the hope that some substitute control method eventually would be found. So far as we can see now, potential substitute methods lie in four different areas: predator- parasite manipulation, propagation of bacterial or viral diseases,
82
Psyche
[June-September
baiting with attractants, and genetic disruption. In briefly discussing these topics, we should not overlook the possibility that there may exist entirely different modes of attacking the problem that have not yet occurred to anyone.
Predators and parasites. As already mentioned, a number of predaceous, parasitic and parasitoid insects, mainly beetles, flies and wasplike types, have been successfully colonized in the United States after being brought from Europe and Asia. Different ones attack every stage of the moth, from egg through adult, but few of them are strictly specific to the gypsy moth. The efficacy of the parasites is now open to question, since they have obviously not prevented serious outbreaks in areas where they are known to be established. Never- theless, some natural enemies are known to be very effective at high densities of the host, and their value in the absence of possibly disturb- ing chemical control has not been thoroughly checked in recent years. Furthermore, it is likely that the established introductions represent only a fraction of the potentially useful arthropod enemies of the moth existing in Eurasia or elsewhere. In theory at least, there remains the possibility of keeping the moth at a tolerable population level by means of natural enemies, especially if used in conjunction with other biological control methods. Further research on natural enemies of the moth would certainly be desirable.
Disease propagation. The gypsy moth larva is susceptible to certain bacterial and viral diseases, among which Bacillus thuringiensis shows enough promise to have stimulated large-scale tests by Federal and state agencies. These tests, only partly completed, employ a “sticker” of tung oil or one of the improved English Eovol products to fasten the bacterial spores to the foliage. The suspension of spores in sticker can be sprayed from the air, and presumably is not harmful to plants or wildlife. So far, results have not been encouraging.
Attractants. The female gypsy moth, as already stated, can flutter along the ground or over low plants, but she cannot truly fly for any distance. The strong-flying males, like those of many moths, are strongly activated, even over long distances, by scent released by the female from the terminal segments or “tip” of her abdomen. Upon sensing even minute amounts of this scent, the male responds by flying upwind, in this way automatically approaching the scent-producing female, and ultimately coming near enough to mate with her. The scent obtained by extracting the female tips in benzol has been used for years as a lure in metal or paper traps to survey suspected areas in order to determine whether males, and therefore a likely infestation, are present. The female tips are obtained by the laborious and extremely expensive rearing of thousands of hand-collected female
1961]
Insect Control Programs
83
pupae, many of them imported from Europe and North Africa. Costs have ranged up to a half dollar per tip in poor collecting years.
In i960, after producing several moderately effective synthetic lures, M. Jacobson and his co-workers of the Entomology Research Division, Agricultural Research Service, USDA, succeeded in isolating the principal sex attractant from some half a million female gypsy moth tips collected in Connecticut and Spain. The substance was prepared synthetically and found to be an ester alcohol with 16 carbon atoms in its main chain. In the course of preparing the natural lure, a closely related substance (with 18 carbon atoms in its main chain) was also found to act as a strong gypsy moth lure.17 This preparation, named gypl-ure , has the advantage that it can be synthesized cheaply and in quantity from ricinoleic acid, a common component of castor oil. Tested in field traps, quantities of this substance as small as one microgram proved equal in luring power to traps baited with the natural lure. In 1961, as this is written, field trials are being carried out to test the efficacy of gyplure-toxicant combination baits in re- ducing moth populations. Included in this program; are “confusion” tests with saturated levels of gyplure in granular and spray formula- tions. Initial technical difficulties have been met, but it is hoped that these can be cleared up during the 1962 season. It will be appreciated that many hopes ride on these crucial trials.
Genetic methods. The success of the screwworm eradication pro- gram (see below) has raised the possibility that the release of sterilized males might be used to control or eradicate gypsy moth populations. This possibility remains to be explored by further studies of the moths’ mating behavior and physiology and the practicability of rearing, sterilization and release procedures. Sterile male release might be made much more effective after reduction of the population by bait attractants or other means.
Other theoretical possibilities for control rest in the fact, discovered years ago by R. B. Goldschmidt, that certain different native Old World populations of P. dispar differ in their sex-determining mech- anisms in such a way that crosses made between them produce inter- sexes. It can be argued that the overall fitness of a population might be cut by introducing north Japanese strains into the American populations, which originated in France. The possibility is worth investigation despite some theoretical difficulties.
THE IMPORTED FIRE ANT Introduction
The fire ants belong to seven or eight New World species in the gerninata group of genus Solenopsis. The group as a whole has a
84
Psyche
[June-September
tropical warm temperate distribution throughout the Americas, from southeastern and southwestern U. S. to central Argentina and Chile. The species are quite closely related and are similar in their habits. All form populous nests, at maturity containing 25,000 to more than 200,000 active and aggressive adult workers. The workers in a mature nest vary considerably in size from large soldiers down to much more numerous minor workers only 2-3 mm, long, and usually only a single functional queen is present. Nest foundation follows the pattern typical for ants, in which virgin winged females mate with males during a nuptial flight, then quickly shed their wings and, as young queens, burrow into the soil and begin the rearing of the first brood in a small chamber. Later, as the nest grows, it usually comes to be capped by an earthen mound sometimes two feet or more high and often two or three feet in diameter.
Up to the First World War, only three of the fire ant species were known to occur in the U. S., of which two, Solenopsis xyloni and S. geminata (native fire ant) were found in the southeastern states. It seems possible that the “native” fire ant is itself a post-Columbian introduction, and it has been spread widely over the tropics of both hemispheres by human commerce. In past years, S. geminata had gathered to itself much the same reputation as a nuisance now gen- erally assigned to the late-coming imported fire ant ( S . saevissuna) that is the subject of this discussion. The imported fire ant arrived at Mobile, Alabama in produce or ballast at or a few years after the end of the First World War. At first the ant (then represented solely, so it seems, by a blackish phase with a dull orange band at the base of its gaster SB- the so-called “variety richteri ” common in Argentina and Uruguay) spread only very slowly in Mobile and its environs. At some time around the beginning of the 1930’s, a smaller, light reddish form of saevissuna appeared in the Mobile area. This phase corresponds to populations of the species common in southern Brazil and Paraguay, and it seems most likely that its appearance marks a second introduction of saevissuna into the Mobile Bay port area.
Coincident with the advent of the red phase, the entire saevissuna salient in southern Alabama entered upon a period of rapid expansion that carried the main infestation across state lines by 1940. The expansion apparently has not yet reached its full extent, although infestations are or have been known to occur in ten states ranging from Texas and Arkansas to North Carolina and Florida. Expansion occurs in two main ways — by steady widening of the main infested areas due to short-range aerial spread of winged females, and through
1961]
Insect Control Programs
85
colonization ahead of the main infested area by queens and colony fragments transported by vehicular traffic. Nursery stock used to be a prime source of new infestations, but since nursery treatments and quarantine regulations have come into effect, fertilized females acci- dently carried in automobiles are probably responsible for most colonization.
Wherever the red phase has expanded to overcome the dark phase, the two extreme forms have interbred to produce a series of inter- mediates, and in most cases the red form soon comes to predominate by a process of genetic swamping coupled with its greater success in warfare between nests. In fact, it may not be too extravagant a speculation to conclude that it was the injection of the red-form genes into the existing dark population that sparked the spectacular spread of the species in the last three decades. At present, the North Ameri- can population consists mainly of light reddish ants, the dark phase surviving mainly in peripheral situations and cool swamplands.
Wherever it spreads, S. saevissima tends to replace the populations of S. xyloni and S. geminata in its path, though this is less true of the dark-colored geminata occupying woodlands in Florida and per- haps elsewhere26; saevissima in the U.S. generally avoids shaded situa- tions. The imported fire ant is able to build up remarkably dense populations. I have seen pastures in eastern Mississippi in which it was literally possible to walk for a considerable distance by stepping from mound to mound without touching a foot to the ground between. Such situations are exceptional, and usually mark the entry of the species into a new area, or else follow control measures that have knocked out a stable population of old, large nests. When the old nests are eliminated, large numbers (up to 185 per acre) of smaller new ones take their places, but as they grow, nests are gradually eliminated until the density is again relatively low (10-50 nests per acre usually).
Studies made to date have not been critical enough to detect possible widespread population fluctuations in untreated areas, but about a century ago, Bates noted a radical change in a native population of S. saevissima in the Amazon Basin.
A small number of parasites of this ant are known in its native habitat, including several known or suspected inquilinous species of ants and a phorid fly, but no real study has ever been made of this phase of the ant’s biology. These parasites have been lightly dismissed as a control possibility by previous writers, but it seems to me that the whole subject of parasitism should be looked into. Parasites might do
86
Psyche
[June-September
much better in the U. S. than in their native range, and even a minor reduction in fire ant populations might reduce it appreciably as a nuisance in some areas.
Nature and Extent of Damage
The kind and extent of the damage done by fire ants has been the subject of much dispute. Generally, control agencies, and especially the USDA-affiliated ones, have emphasized the deleterious effects produced by the1 ant, while some zealous anti-insecticide writers have written it off as doing negligible harm. Both groups admit that the ant mounds do interfere with the harvesting of forage crops. Harvest- ing machinery is often damaged by striking the hard mounds, and field hands are stung by the ants — in some cases so badly that they refuse to work infested fields. Occasionally, land values have fallen somewhat in badly infested areas. The health threat must also be considered in cities and towns, where the ants may infest lawns and gardens and even sometimes enter houses. Small children and unusu- ally sensitive adults have occasionally suffered grave illness, or in two or three cases may even have died as a result of fire ant stings. Numer- ous stings result in a rash-like group of pustules that can be very an- noying for several days or more. Still, the fire ant as a health menace must be ranked far below ordinary bees and wasps, which are respon- sible for many times the deaths that fire ants cause during a given period of years, in the same states. It is difficult to see how the ant can be classed as a serious public health problem despite scare stories in the press, television and in a USDA-sponsored film. Professor F. S. Arant, head of the entomological contingent at Auburn University, current president of the Entomological Society of America, and a top authority on the fire ant, agreeing with Dr. J. L. George10 and other state entomologists in the Southeast, calls the fire ant a “major nuisance,” but deprecates its role as a crop pest. Studies made at Auburn14 and elewhere in the South generally have borne out this evaluation. It is interesting to note that the studies6’27 that have found more or less serious damage done to crop plants were made before 1953. These studies were mainly concentrated in south-central Alabama, near the Mobile Bay center of fire ant spread, and were based on personal investigation as well as uninvestigated farmer reports. That some crop damage was done in this area in the late ’forties and early ’fifties is incontestable, but even then, the damage does not seem to have been insupportable. That more recent studies have failed to find serious crop damage is probably to be laid to a gradual change in the habits of the ants or their population density,
1961]
Insect Control Programs
87
or both. Whatever is the case, it does seem that the damage currently being done by the imported fire ant in the untreated sections infested in this country is less than would seem to justify the massive campaign that has been mounted against it. Agencies in all but two infested states do not even grant the fire ant a place in their lists of the more important plant pests. The USDA cites farmer support for the program, and this support certainly exists at least in some sections. But the enthusiasm of farmers for the spray programs is too often based merely on a vague feeling that insecticides in general are a good thing. When, as in large areas covered by the present program, the farmers individually get the spray free, they tend to overlook possible bad effects it may bring with the benefits. In any case, the satisfaction of farmers is certainly no substitute for a careful and extensive professional check of current fire ant damage. No such check has been made by the USDA, or at least none has been reported upon since 1952.
Control Operations
Control efforts directed against the imported fire ant were first initiated on a small scale by the State of Mississippi in 1948, without notable success. A survey of the infested area was begun by the USDA in the fall of 1948, and, together with limited investigation of the biology of the ant and control measures against it,6 ran until research funds were stopped in 1953. This investigation did not deal with aerial control measures, and little attention was paid to wildlife damage. It is important to note that from 1953 until 1958, after the USDA had started its mass spray program, it spent no money for fire ant research.22 Meanwhile, several independent agencies had done part-time research on various aspects of fire ant biology and control, including medical studies of the effects of the venom on humans at Tulane University, biological and control studies at Auburn and Mississippi State Universities, and behavioral and other investigations by Dr. E. O. Wilson and others (including the present author) at Harvard University and in the field. The Fish and Wildlife Service, although greatly hampered by lack of research funds for this purpose, was giving some attention to the prospect of mass broadcasting of insecticides as it could be expected to affect wildlife.
Against this patchy research background, in March, 1957, the USDA noted that it had requested the approval of Congress for control of the fire ant, and Congress forthwith passed a special “Fed- eral Plant Pest Act,” authorizing the USDA to take measures against
88
Psyche
[June-September
the ant. For the 12 months beginning July, 1957, 2.4 million dollars was appropriated, to be matched by funds from state agencies, local sources and/or individual farmers. (In practice, actual matching appears to have been spotty at best, and the government has waived farmer contributions in Georgia and parts of Florida since early in the program.)
On April 18, 1957, after a brief correspondence with officers in the Entomology Research Division of the Agricultural Research Service, USDA, I received a letter from Dr. A. W. Lindquist, head of one of the sections in the Division, which started in part as follows:22
“The idea of airplane spraying and dusting for control probably stems from the fact that extensive areas are infested. This method of application would of course be fine if it were effective. However, we would want to see considerable research conducted to determine if it would be effective and, if so, to determine what insecticides and special precautions would be necessary for maximum results. As far as we know, no research along these lines has been conducted.”
This answer may be compared with that received from Dr. M. R. Clarkson,23 Acting Administator of the Agricultural Research Serv- ice, dated January 3, 1958, stating in part:
“In planning field operations, all available results of applicable research and practical experience are taken into account. Close liaison has been established with the Fish and Wildlife Service of the Depart- ment of the Interior and the states involved. Competent wildlife observers have been assigned to the work and experience to date indicates that a successful program can be carried out without serious consequence to wildlife resources. . . . Both the Agricultural Research Service and State Experiment Stations have expanded their research program in a continuing effort to improve operational procedures.” ( Italics mine — W.L.B. )
In May, 1957, as a matter of record, Dr. Ross Leffler of the Department of the Interior had written to Representative H. C. Bonner, Chairman of the House Committee considering the bill, as follows in part :
“Sufficient basic research has not been accomplished to predict losses or to properly advise operating agencies on the means of obtaining effective control and at the same time avoiding unnecessary fish and wildlife mortality.”
With astonishing swiftness, and over the mounting protests of con- servation and other groups alarmed at the prospect of another airborne “spray” program, the first insecticides were laid down in November, l957- The rate of application was two pounds of dieldrin or heptach-
1961]
Insect Control Programs
89
lor per acre, the insecticides being incorporated in granules of an inert material to cut down wind drift and lessen loss by foliage interception. It had been established that this formulation would be spread in the upper soil layers when rain dissolved the granules, and that its effect would last at least three years.1 Dieldrin was used at three pounds per acre wherever another pest, the white fringe beetle, occurred as well as the ant, thus treating for both pests at once. Where the ant occurred alone, heptachlor was usually the choice. Dieldrin and heptachlor are extremely toxic substances — about 4-15 times as toxic to wildlife as is DDT.8 Many wildlife experts and conservationists, as well as entomologists both basic and economic, felt a sense of foreboding at the start of a program that would deposit poisons with 8-30 times the killing power of the common forest dosage of DDT (one pound per acre in gypsy moth control).
The spray campaign got off to such a fast start that both state and Federal agencies were caught without being able properly to organize programs that year for assessing the effects of the poisons on wildlife, so that results of such programs were delayed until after large amounts of toxicants had already been laid down.
Now that some of these results are finally available, we can see that they were acutely needed before the program was ever begun. The misgivings of the wildlife people seem to have been justified on the whole, since the kill of wildlife in sample treated areas appears to have been high in most of those that were adequately checked.5’ 8> 10, 12> 21 The USDA disputes many of the claims of damage, but their own statements often tend to be vague and general. It does seem to be true that quail and perhaps other wildlife species will make a good come- back on treated land after two or three years, provided that untreated areas are available nearby to furnish replenishment stocks once the treated land begins to recover. Still, most of the information on wild- life repopulation comes from the accounts of hunters and other sources not subject to proper checking, and we still have little in the way of published studies by competent authorities on ecological recovery of treated lands.
Wash-off into streams and inlets has led to heavy losses among fish, crayfish and aquatic insects. Dieldrin at only one pound per acre sprayed on a salt marsh at Vero Beach, Florida, killed all the fish (including young tarpon) and Crustacea in the marsh and adjacent waters, and the effect lasting for weeks.12 This particular test, meant to control sandfly populations, applied only half of the dosage of dieldrin originally used for fire ant control, and one-third the dosage actually used on white fringe beetle together with fire ant.
90
Psyche
[June-September
Although the USDA claims that the evidence is inconclusive in some cases, there does exist contrary information7, 10 indicating that stock losses from fire ant poisons may sometimes be significant. Various newspaper accounts, while sensational in tone and possibly exag- gerated, add further to the impression that damage to cattle, horses, poultry and household pets may on several occasions have been locally serious. Even a few livestock deaths, if added to the time and effort spent by farmers in carrying out awkward measures to protect their animals from spray measures, must more than balance out any cumu- lative loss that fire ants may have inflicted directly on farm stock since the infestation began.
In 1959, the formulation was changed to a dosage of 1.25 lb of dieldrin or heptachlor per acre, and more recently an alternative dosage of a quarter pound per acre has been most widely used. This latter dosage, used twice at three- to six-month intervals, was devel- oped because of the growing concern about wildlife and the residue problem. At this rate of application, wildlife apparently suffers much less seriously, but the fire ant is also much safer than under the old rate of two pounds per acre, and can probably come back in many places a year or two after the “light treatment” has been applied, according to the data of Blake, Eden and Hays1 for similar dosages. Wildlife officials claim to have heard from Plant Pest Control officers that there still exist stockpiles of the formulation yielding two pounds of actual heptachlor or dieldrin per acre, and that this product was still being used for treating junkyards as of March, 1961, but Dr. E. D. Burgess of Plant Pest Control denies that this is so.
A serious blow was dealt the program in late 1958, when treat- ments were only one year old ; Senator Sparkman and Congressman Boykin of Alabama asked that the fire ant campaign be suspended until its benefits and dangers could be evaluated properly. Then, in the beginning of i960, the Food and Drug Administration of the Department of Health, Education and Welfare lowered the tolerance for heptachlor residues on harvested crops to zero, following the discovery that heptachlor was transformed by weathering into a per- sistent and highly toxic derivative, heptachlor epoxide, residues of